Bulletin OF THE Illinois State Laboratory OF Natural History Urbana, Illinois, U. S. A. Vol. VIII. Article I. THE PLANKTON OF THE ILLINOIS RIVER, 1894-1899, WITH INTRODUCTORY NOTES UPON THE HYDROGRAPHY OF THE ILLINOIS RIVER AND ]TS BASIN. PART II. CONSTITUENT ORGANISMS AND THEIR SEASONAL DISTRIBUTION. BY C. A. KOFOID, Ph.D. ERRATA AND ADDENDA. Page 58, line 7, for ovalis read ovata. Page 85, line 8, for longicaiidiis read longicanda, and just above Phaciis plcitro- nectes read the following paragraph : — Phacus longicanda var. torta. n. var.—This variety, for which I propose the name torta because of the twisted body, is figured by Stein ('78, Taf. 20, Fig. 3). It occurred sparingly in midsummer from July to September, rarely in October, in 1896 and 1897. Page 91, line 18, after T. caudata Ehrb. read T. lagenella Stein. Pages 153, line 3 from bottom, 168, line 16, and 178, line 14, iov'98 read '.9Sa. Pages 156, line 11, 159, line 16, and 161, line 5 from bottom, for '93 read '98a. Pages 175, line 5, 186, line 3, and 208, line 17, for Bimcerium read Dimcerium. Page 288, line, 3 for Lampsilus read Lampsilis. Page 292, line 13, for gracilis read gracile. Page 471, line 3 under heading beetles, for pennsylvanicus read pennsylvanica VI PAGE Coelenterata 132 Platyhelminthes 133 Turbellaria 133 Trematoda 134 Cestoda 135 Nemertini 135 Nematelminthes 135 Nematoda 135 Acanthocephala 136 Annulata 136 Oligochseta 136 Naidida; 137 ^olosomatidffi 137 Rotifera 138 Rhizota 139 Discussion of Species 140 Bdelloida 141 Discussion of Species 141 Ploima 144 Pulses of Ploima (Table) 146 Discussion of Species 147 Tabular statistical data : — Pulses of AnurcBa cochlearis 150, 153 Pulses of A nurcBa hypelasma 155 Pulses of Asplanchna brightwellii 156 Evidence of polycyclic character of seasonal distribution of above . 158 Evidence of same for Asplanchna priodonta 161 Pulses of Brachionus angularis 165 Collection data for Brachionus hakeri and varieties 169, 170 Pulses of Brachionus budapestinensis 176 Pulses of Brachionus militaris 179 Pulses of Brachiomis mollis 181 Brachionus pala and varieties. Average number per collection . 182 Pulses of Brachionus pala and varieties 184 Brachionus pala, type form. Sexual cycle 188 Seasonal distribution of B. pala and its var. amphiceros 189 Seasonal limits of B. pala var. dorcas 191 Pulses of Brachionus urceolaris 193 Brachionus urceolaris and varieties and B. variabilis. Average number per collection 196 Pulses of Polyarthra platyptera 204 Pulses of Synchccta pectinata 211 Pulses of SynchcBta stylata 213 Pulses of Triarthra longiseta 218 Scirtopoda 219 Seasonal fluctuations of Pedalion mirum (Table) 220 Gastrotricha 221 vn PAGE Entomostraca • 221 Branchiopoda 222 Cladocera 223 Cladocera and Hydrographic Fluctuations (Table) 223 Discussion of Species 225 Tabular statistical data: — Bosmina and hydrographic fluctuations 228 Seasonal distribution of Chydorus. Average number per m.^. . . 235 Effect of temperature on numbers of Chydorus 238 A typical pulse of DapJinia cucnllata 242 Diaphanosoma population, with data of temperature and river level 248 Moina and hj'^drographic changes 254 Ostracoda 257 Discussion of Species , 258 Copepoda 258 Copepoda and Hydrographic Changes (Table) 260 Discussion of Species 261 Cyclops viridis var. insectiis and Hydrographic Changes (Table). . 274, 275 Relative numbers of adult, yotxng, and larval Cyclopidag (Table). . 278 Amphipoda 283 Arachnida 283 Acarina 283 Tardigrada 284 Hexapoda 284 Ephemerida 285 Hemiptera 285 Diptera 285 Mollusca 287 Gastropoda 287 Lamellibranchiata 287 Bryozoa 288 Discussion of Species 288 The Periodicity in the Multiplication of the Organisms of the Plankton. . . 291-311 Comparison of Plankton Pulses and Lunar Cycle (Table) 296-299 Gaseous Contents of Pond Water (Table) 306 Relation of Pulses of Chlorophyll-bearing Organisms to Lunar Cycle (Table) 308 General Considerations on Seasonal Changes 312 Lake versus River Plankton 312 Organisms per Cubic Meter in Plankton of Illinois River in 1898 (Table) . . 314-*340 Bibliography 341-354 Explanation of Plates 355 Diagrams of Seasonal Distribution 356-360 Errata and Addenda 361 Article I. — Plankton Studies: V} The Plankton of the Illi- nois River, 1894-1899. Part II. Constituent Organisms and their Seasonal Distribution. By C. A. Kofoid. Introduction. This paper gives the results of a statistical study of a series of quantitative plankton collections made in the channel of the Illinois River near Havana, 111., at the Illinois Biological Station, in 1894- 1899. The environmental conditions and the volumetric results of this investigation have been given in Part I. (Kofoid, '03), published in Volume VI. of this Bulletin. Of the 235 collections made in channel waters and used in the quantitative study, only 182 were subjected to numerical and qualitative analysis. The omitted collections were intercalated at brief intervals of one to several days between those enumerated, principally in the summer of 1895 and during the winter flood of 1896 and the summer of the same year. The collections chosen for this study, whenever possible, represent a weekly interval, and a full list of all collections, with environmental data, may be found in Table III. of Part I. The chronological distribution of the collec- tions studied by the statistical method is given in the table on the following page. The work of enumeration and the primary tabulation was com- pleted at Urbana December 31, 1900, when my formal connection with the State Laboratory ceased. The manuscript has been 1 The four preceding numbers of this series, all by the present writer, have been published as articles of the Bulletin of the Illinois State Laboratory of Natural History, as follows: — Article I., Vol. V.—Plankton vStudies. I. Methods and Apparatus in Use in Plankton Investigations at the Biological Experiment Station of the University of Illinois. Article V., Vol. V.—Plankton Studies. II. On Pleodorina illinoisensis , a New- Species from the Plankton of the Illinois River. Article IX., Vol. V.—Plankton Studies. III. On Platydorina, a New Genus of the Family Volvocidce, from the Plankton of the Illinois River. Article II., Vol, VI.—Plankton Studies. IV. The Plankton of the Illinois River, 1894-1899, with Introductory Notes upon the Hydrography of the Illinois River and its Basin. Part I. Quantitative Investigations and General Results. (2) prepared at Berkeley, being completed in May, 1904, after my connection with the University of California was begun. My separation from the collections and the library of the State Labora- tory has rendered impossible some verifications, comparisons of specimens with more recent literature, especially among the algas. Distribution of Collections by Months. METHODS. The collections were preserved in bottles of uniform capacity (60 cm.^), in alcohol-formalin mixture (2 per cent, formalin in 70 per cent, alcohol), and after measurement by the centrifuge were released from the compressed condition in the measuring tubes and returned to the containers. The counting was done by a modified Sedgwick-Rafter method (see Kofoid, '97), in which 1 cm.-'' of a suitably diluted plankton is distributed evenly in a cell 20 x 50 mm. The plankton was diluted or condensed (from 60 cm.^ of fluid) according to the quantity of plankton and the amount and nature of the silt. Larger organisms such as the Entomostraca were counted in the whole catch, or in larger collections in Vio to Vso of the total catch ; and the smaller organisms in V25 to V^o- The filter-paper catches which supple- mented those of the plankton net from August 3, 1896, to the end of the series, March 28, 1899, were often subjected to considerable dilution on account of the great amount of fine silt in the collections, from Vio fo Vioo being the limits of dilution as a rule. The even distribution of the organisms in the Rafter cell was secured by shaking the collection in a mixing cylinder gently till the sediment was thoroughly distributed, and taking the sample immediately with a long 1 cm.^ pipette, inserted to the bottom of the jar and raised to the surface during the filling process, and by discharging the contents immediately into the cell at one corner, the cover having been previously displaced at a slight obliquity to admit the end of the pipette. With the filling of the cell the cover automatically moves into place, and practice soon enables one to fill the cell without inclusion of air bubbles. With the exception of the heavier rhizopods, all of the organisms are as a rule very evenly distributed by this method. The identification and enumeration of the contents of the cell were carried on with the help of a mechanical stage and a § Bausch & Lomb objective, with a Zeiss C for higher magnification when needed for the detection of fine details or for counting the smaller organisms in the filter-paper catches. After considerable experimenting, the following method was established in the work of enumeration. Four sheets, each with numbers 1 to 76 at the left, were fastened temporarily to accom- panying key sheets, each number on each sheet standing for one of the more common species. One sheet was assigned to algae, diatoms, and miscellaneous organisms; and one each to Protozoa, Rotifera, and Entomostraca. As the plankton sample was examined under the microscope the identifications were called off, and entered on the sheets by a clerical assistant. Six of the most abundant species were recorded by the observer himself on six tallying machines registering 1,000, and conveniently arranged in a box at his right. By adjusting the springs to give different sounds when registry was made, and by modifying the surfaces pressed by the fingers so as to differentiate the several machines without looking at them, it was possible to use these without raising the eye from the microscope, and thus to avoid the fatigue arising from the repeated muscular readjustment of the eyes necessary when the observer makes his own entries in a written record. Common species not recorded by the tallying machines were generally abbre- viated or designated by easily-called tokens. When once fairly familiar with the species it was possible by means of these labor- saving devices to make identifications and enumerations of several heavy planktons per day. By a number of tests I found that when the enumerations of a species in a given collection reached 1,000, little was gained by carrying it to higher numbers. A limit of error of +5 per cent, can be thus obtained if the species in question is distributed evenly in the cell and all precautions are observed to secure accuracy. Enumerations were often carried beyond this point, but rarelybeyond 3,000. The accessions numbers of the collections from our catalog of collections served to designate each sheet of data and all note slips bearing on the collection or its constituent organisms. When the enumeration was completed, the factors of collection, dilution, and enumeration were entered on the sheets, and the number of individuals of all species represented was computed and carried to the right of the sheet. The totals of the various groups—for ex- ample, diatoms or Cladocera—were then added up and entered on the sheets in differential colors. By the use of the key sheets the number per m.^ of water of any given species could be quickly ascer- tained. Species not in the key were entered by name on the sheets. When the enumeration of all collections was completed, the numbers per m.^ giving the seasonal distribution of the various species and groups through the collections of 1894-1899 were drawn up on uniform folio sheets, and the annual totals and averages com- puted therefrom. With the data in these forms it is possible to turn at once to the statistics of the plankton of a given day, or to the seasonal distribution of any desired species. ACKNOWLEDGMENTS. I am indebted to Prof. S. A. Forbes, Messrs. E. B. Forbes, F. W. Schacht, and R. W. Sharpe for many suggestions concerning the Entomostraca; to Prof. Frank Smith for assistance with the Oligo- chccta of the plankton ; and to Mr. A. Hempel for my introduction to the Rotifera. The identification of the cosmopolitan species of the fresh-water plankton of the Illinois River was greatly facilitated by the most excellent library of the Illinois State Laboratory of Natu- ral History, the accumulation of many years' careful selection by its director. Prof. S. A. Forbes. The literature of fresh-water fauna, and to a large extent of its flora also, is very fully represented therein. The excellent Laboratory collection of identified Entomostraca from European specialists was also of great service. I am indebted to Mr. R. E. Richardson for valuable services as clerical assistant, and for substantial help in organizing the great mass of data resulting from the enumerations. Except as noted in the discussion in subsequent pages, I hold myself responsible for all of the identifications of the species re- corded. The enumeration is also all my own work, with the excep- tion of that of the nauplii, of two species of Difflngia, and of Pedi- astrum in about one third of the collections, in which I had the assistance of Mr. R. J. DeMotte, and that of the commoner Rotifera in a few of the collections, which were counted by Mr. Richardson. DEFINITIONS. The term "plankton" was used by Hensen ('87) to designate "Alles was im Wasser treibt." It w^as applied by him only to that assemblage of marine organisms which float passively in the open sea, without active recourse to shore or bottom, and unable by their own efforts materially to change their location. The term has since been extended also to assemblages of organisms in fresh water which bear a similar relation to open water. This fresh-water 8 plankton has been designated in turn " limnoplankton " by Haeckel ('90), a word which in a restricted sense is retained for the plankton of lakes, while that of rivers has been distinguished by Zacharias ('98a) as "potamoplankton," and that of ponds ('98) as "heleo- plankton." These distinctions are based upon the nature of the en\'ironing body of water, and the terms are convenient, though the separation of these types everywhere in nature is difficult, if not impossible. Owing to the smaller size of fresh-water basins as compared with those of marine character, the shore and bottom be- come more important as factors in the environment of the plankton. Within the fresh-water environment we also find degrees of impor- tance of the shore and bottom which in ascending scale dominate in the lake, river, pond, and marsh. Although each of these repre- sents distinct conceptions, in nature we find them imperceptibly intergrading, and neither these conceptions, geographical nomen- clature, nor local parlance give us any final criterion which will enable us to use the terms with the precision which a scientific terminology would demand. The distinctions between these forms of fresh-water plankton must lie in the plankton itself, if anywhere. As I shall attempt to show later, these distinctions, though appar- ent, in some cases at least, are nevertheless of minor importance, and depend very largely upon the relative predominance of the adventitious littoral fauna and flora rather than upon distinctive assemblages of eulimnetic species. The striking similarity of this eulimnetic plankton in all these types of environment and in widely separated continents is a biological phenomenon of far more sig- nificance than these minor difi'erences. These distinctions between the different types of fresh-water plankton are thus more a matter of terminology than of biological import. Among the organisms found in open water there are varying degrees of dependence upon the shore and bottom. Some, as Cyclops and many of the lower algas, have life cycles in which no encysted or quiescent resting stage has been found, and actively or passively their whole existence is passed in the open water. They are at all times components of the plankton ; that is, are continuous planktonts. Others, as Dinohryon, many of the Rotifera and Cladoc- era, and, in fact, the greater part of the eulimnetic organisms, have an encysted stage which as a winter egg or a cyst descends to the bottom and remains there for a season. Such organisms only periodically, wholly or in part, leave the open water for a littoral or benthal existence. They are periodic planktonts. Some organ- isms, such as many of the rhizopods and diatoms and Hydra, appear in the plankton under certain conditions of temperature and food. They temporarily adopt the limnetic mode of life as a result either of a change in their specific gravity due to internal changes, such as an increase of the gaseous or fatty contents of their protoplasm, or to changes in the buoyancy of the water due to changes in temperature or in substances in solution in the water, or because of the abundance of food in the open water. They become under these conditions actively adventitious planktonts. Still other organisms are released from their usual contact with or attach- ment to the substratum, or from their association with debris or vegetation of shore or bottom, by movements or disturbances in the water, and are swept into the open water only to return again to their customary habitat when conditions favor. Practically all of the smaller organisms inhabiting the shore and bottom and the debris and vegetation found thereon are liable thus to enter the open water, and to be found in forced and temporary association with the eulimnetic fauna and flora. They are passively adventi- tious planktonts . Another class of organisms which occur in the plankton are those which either as internal or external parasites find in plankton organ- isms either a host or a substratum for attachment. These are in a certain sense passive planktonts, and they may be distinguished from other passive planktonts as attached or parasitic planktonts. Sharp lines betw^een these various classes of organisms found in open water can not be drawn upon distinctions based upon their degree of dependence upon the bottom and shore. An equally vague line separates the organisms of the plankton from those more active forms which by virtue of their powers of locomotion are to a con- siderable degree independent of waves and current, and are able freely to maintain their position in their preferred habitat. - Among the organisms commonly included in the plankton, the flagellates, rotifers, and Entomostraca exhibit some degree of activity, such as is seen in their limited vertical migrations, while larger organisms, such as Leptodora hyalina and the larvas of Corethra, are capable of movement sufficient to give them considerable independence in the matter of their position in the water. We thus find degrees of inde- 10 pendence which approach closely that found in young fish and the large insect larvae—organisms not always regarded as planktonts. The plankton is thus a composite assemblage of organisms whose association depends in varying degrees upon their relation to their common habitat, the open water. In actual practice, all the organ- isms found in the open water are regarded as within the scope of plankton investigations, and justly so, for by virtue of their pres- ence they become more or less involved in the complex interrela- tions which pertain to the flux of matter, the succession of species, and the food relations which exist through the changing seasons in the aquatic environment. In our own investigations it has been our purpose to include all the organisms found in our collections; that is, all which our meth- ods of examination give us a sufficient means of investigating. Naturally, the bacteria are to large extent excluded from our con- sideration, though they properly belong to the plankton, and in the processes of nitrification and , denitrification play an exceedingly im- portant part in the economy of aquatic life. THE COMPOSITION OF THE BLANKTON. The composite character of the plankton is especially marked in streams,—as, for example, inthe Illinois River,—owing tothemingling of organisms from a great variety of tributary sources—backwaters, lakes, ponds, pools, marshes, swamps, brooks, rivers, canals, sewers, drains, and industrial wastes. Few lakes possess so varied a supply, and in none can the proportional effect of these contributions exceed that of the stream. Added to this contributed assemblage, and in some seasons predominating over it, is the indigenous or autono- mous plankton of the stream itself. The component organisms of the plankton of the Illinois River number 528 forms, including only those which have been identified from collections made in the main stream and including both species and well-defined forms or varieties. Species found thus far only in the backwaters are not included, though there is little doubt that they occur also in the main stream. No effort has been made to build up merely a long list of species, but only to identify, so far as possible, the common and recurring forms. Neither has any attempt been made to establish new species or revise those already 11 described, though a magnificent opportunity awaits the naturaHst who has the fortitude to analyze the exceedingly variable forms which compose the plankton, and to determine by modern methods which of these variants are entitled to specific rank. It has seemed to the writer that the only satisfactory basis upon which species, and pre-eminently those of the fresh-water plankton, can rest, lies in a careful determination of the limits of seasonal and local variation within the area of distribution. This means breeding under con- trol, and the study of variation by modern statistical methods. Both of these lines of inquiry lie beyond the purpose of the present paper, and plainly beyond the possibilities of accomplishment by any one investigator, when the great number of species and the pres- ent state of the literature of the subject is considered. It is becom- ing constantly more evident that the species of the plankton are in the main cosmopolites, and the world literature of the subject must be taken into consideration in any thorough attempt to handle the systematic side of the subject. During the progress of this work, which was begun in 1894, every effort was made to secure all perti- nent literature bearing on the genera of plants and animals repre- sented in the plankton, and so far as possible in the enumeration of the collections the individuals were referred to "species" already described, or, in default of this, recorded as "unidentified." In some groups—notably the desmids, diatoms, and unicellular alg«—it was not possible under the conditions of plankton enumer- ation to apply to all the individuals enumerated the fine distinc- tions which specialists in these groups have made. They have been throw^n under certain of the better-defined species, which thus stand in our records as representatives of closely related variants as well as of the types of the species named. Examples of this appear in Closteriimi, where two species only were listed. Probably a num- ber of so-called species among the scores described in this genus will be found among the individuals in our plankton here referred to the two species C. acerosum and C. lunula. So, also, in the case of Melosira; tw^o principal types were listed, M. varians and M. granulata,—though even these two seem at times to intergrade. Other described species will be found among the individuals thus distributed. In the case of Difflugia glohulosa and D. lohostoma a large number of intergrading and variable forms are included. It would be possible to find among these, representatives of many 12 recently described species. In these instances the difficulty lies not so much in finding representatives of these closely related species, but, rather, in drawing the lines between them and placing every individual enumerated in the proper pigeon-hole. To avoid this difficulty, the separation was not attempted in every case. With the hope that the results would throw some light on the ques- tion of seasonal variation, this separation was attempted in the genus Brachionus, where the species characters are confined to prominent structural features. So far as it was feasible, specific distinctions were accepted as found, and utilized whenever possible. In the lists and discussions which follow, the inclusion of a species does not necessarily carry with it the inference that it is regarded by the writer as valid or well founded. It merely represents in our enumerations a more or less continuous succession of organisms which conform approxi- mately to the descriptions and figures of the species designated by the name in question. Inferences regarding the rank or validity of the species reported will be given whenever the statistical data or my observations on the variability of the organism seem to afford data bearing on the standing of the species. While not a few of the species reported may justly be regarded as synonyms, an effort has been made to use only names which represent valid species or at least a variety or a seasonal form. COMPARISON OF FRESH-WATER AND MARINE PLANKTON. The plankton of fresh water is very generally composed of an assemblage of organisms, of plants and animals, principally crypto- gams and invertebrates. Not all orders are represented, and those that do occur vary greatly in the number of their representatives. The fresh-water plankton differs from that of the sea in the almost universal absence of larval forms, in the smaller number of inverte- brate groups represented, and in the smaller size of its component organisms. Fresh-water plankton has almost no limnetic coelen- terates. Hydra fusca being the only representative as yet discovered in our locality. The absence of the larger Crustacea, of limnetic mollusks and worms, and of tunicates and Radiolaria robs limnetic life of the diversity found among pelagic organisms of the sea. The only larval stages found in our locality are the glochidia of the 13 UnionidcE, whose limnetic sojourn is at the best but brief, and the larvae of certain dipterous insects, such as Chironomus and Corethra. The limnetic habit of these larvae is hardly established as yet. The small size of fresh-water planktonts as contrasted with those of the sea is very striking. Representatives of the same group—for exam- ple, the Dinojiagellata and the Entoniostraca—in the two habitats exhibit this contrast. The largest entomostracan of fresh water is less than a centimeter in length, and there is nothing to compare with the pelagic coelenterates, Mollitsca, or such tunicates as Salpa and Pyrosoma. The smaller size of fresh-water planktonts may be due to the lower specific gravity of the environing medium, and perhaps also to the effect of smaller quantities of dissolved salts upon the metabolic processes of limnetic animals. Notwithstanding this absence of large individuals in the plank- ton of fresh water, the total quantitative production of plankton per cubic meter is greater here than in the sea. For example, the average production in the Illinois River is 2.71 cm.^, and the average amount in adjacent backwaters rises as high as 22.55 cm.^ (in Phelps Lake). These measurements were made by the centrifuge, and the results of the "Plankton Expedition" of Hensen reduced to this basis of measurement by Kramer ('97) show that the Atlantic Ocean at the time of this expedition had in the upper strata exam- ined but 0.12 to 0.48 cm.^ of plankton per cubic meter. ORGANISMS OF THE PLANKTON. The groups of plants represented in the plankton of the Ilhnois River are principally algae, of which the BacteriacecB are but partially retained in the collections and are usually omitted in plankton investigations. The SchizophycecB, or blue-green algse, furnish a few important representatives and a number of adventitious species. The ChlorophycecE, or green algas, on the other hand, abound both in species and individuals, and afford an element of great impor- tance in the primal food supply. The Bacillariacecr are exceedingly abundant, and are represented by a number of eulimnetic, as well as many adventitious, species. They also constitute one of the primal sources of food for the zooplankton. The ConjugatcB furnish but few species and indi\'iduals—principally desmids—to the phytoplankton. The phanerogams afford a few species which 14 are often taken with the plankton by virtue of their semi-limnetic habit, but do not in the living state enter the food cycle of the plankton nor affect its economy except as competitors. The zooplankton includes representatives of a considerable range of groups, though both in species and individuals the Proto- zoa, Rotifera, and Entomostraca predominate among the animals. Representatives of other groups are in the main adventitious. Among the Protozoa, the Rhizopoda are constantly represented by many individuals and a considerable number of species, many of which may be adventitious, but most of which are wont to adopt the limnetic habit during the warmer months. The Heliozoa are few both in species and individuals. The Mastigophora (which in our discussions include all green and brown flagellates often clas- sified with the ChlorophycecBQ-ndPhcEophycecB) vie with the Chloro- phycecB and BacillariacecB for the first place as converters of the inorganic (and perhaps also the dissolved organic) matter into food for the zooplankton. They are exceedingly numerous in our plank- ton both in species and individuals, and form quantitatively a con- siderable part of the plankton during the summer months. The usual method of plankton collection—by silk bolting-cloth—per- mits a large proportion of these organisms to escape. The Ciliata furnish a few constant members of the plankton, and numerous adventitious and parasitic species. During the low water of autumn, when bacterial contamination is at its height, these organ- isms form a large part of the plankton. The small size of some of the ciliates, combined with their motility and flexibility, renders the loss by their escape through the silk net considerable. The Suctoria furnish but few species and individuals—mainly adventi- tious or attached to other planktonts. The Rotifera constitute, both in species and individuals, the most important single group of analytic organisms, that is those of distinctly animal metabolism, occurring in our plankton. This may in part be due to our shallow warm waters and to the abundance of ChlorophycecB and Mastigophora, w^hich enter largely into their food. This abundance of the Rotifera may prove to be character- istic of the plankton of rivers ( potamoplankton) as contrasted with that of lakes (limnoplankton) . While many rotifers are eulimnetic, the plankton also contains numerous adventitious species. 15 The Entomostraca include the largest fresh-water planktonts, and in every respect constitute an important element of our river plankton. They form the final link in the food cycle which con- nects the nutrients in solution in the water and in decaying detritus with the fish and other aquatic vertebrates. They include numer- ous species, some of which are adventitious. All of the Ostracoda belong to this latter class. The Cladocera furnish some of the most important eulimnetic species and a large number of adventi- tious forms, while the Copepoda are almost wholly eulimnetic. In addition to these groups, the Tiirbellaria, Oligochata, Hexap- oda, Hydrachnida, Gastrotricha, and Bryozoa furnish a few species and individuals of a semi-limnetic or adventitious character to the plankton. In the table which follows, these various groups are listed, and the number of forms occurring in each is noted. In order to give some idea of the proportionate representation of these groups in our plankton, the table includes the sum of the number of indi- viduals per m.^ of water in the weekly collections for the year 1898. This was a year of no marked departure from the normal regimen of hydrographic conditions (Part I., PI. XII.). The summer and autumn flushes tend to lower the population somewhat below that of more stable seasons, but beyond this feature there is nothing to suggest that the plankton of this year may not represent a fair average of that recurring each year in the Illinois River. The fig- ures given, in all cases refer to the number of indi\-iduals per cubic meter (excepting only such cases as Syntira and Uroglena, where the colony rather than the individual becomes the unit) . The algae and Protozoa include many species enumerated in filter-paper col- lections, which accounts for the large numbers in some of the totals. The "'number of forms" listed refers to the total number found in the waters of the river during the period of our operations. Some species not noted in 1898 are therefore included. Unidentified forms are not included in the list of number of species, though the groups here listed to which they belong were known. Some forms referred to genera but not determined as to species are, however, included. This table throws some light upon the ecological relations of the groups composing the plankton, since it gives some clue to their relative numbers, and these condition in a general way the food 16 relations existing between the different groups. The plants are more abundant (and generally smaller) than the animals, outnum- bering them nearly 5 to 1 . Computation shows that for each one of the Cladocera there are 7 Copepoda, the predominance of the latter Constituent Groups of the Annual Plankton of the Illinois River. Average of 52 Weekly Collections in 1898 — number 'per m^. Algas: Bacteriaceas. . Schizophyceae. Chlorophyceae. Bacillariaceffi. Conjugatje. . . . Phanerogainia .... Total phytoplanktonts. Protozoa—total Mastigophora Rhizopoda Heliozoa Sporozoa Ciliata Suctoria Rotifera Entomostraca—total Cladocera Ostracoda Copepoda Miscellaneous Total zooplanktonts Total planktonts enumerated. Synthetic (chlorophyll-bearing). . . . Analytic (non-chlorophyll-bearing) . Number of forms recorded. Number of individuals. 3 9 33 29 7 2 83 (185) 68 59 5 3 45 5 104 (43) 26 4 13 114 446 529 (57,142,822)* 85,909,985 53,175,105 396,192,716 48,459 9 535,326,274 (111,731,000) 95,856,449 55,364 4,871 1,638 15,812,346 332 592,416 (47,041) 6,242 191 40,608 9,393 112,379,850 647,706,124 613,017,986 34,687,781 being accounted for in part by the fact that their larval stages are free-swimming and appear in the enumerations, while the young of the Cladocera are not set free until nearer maturity. About 10 to 20 per cent, of the Copepoda are adults. The relative numbers of * Represents fragments of filaments, and is not included- in totals. 17 the two groups are not so disproportionate as the figures might seem to indicate. For each one of the Cladocera there are 95 roti- fers and almost 18,000 Protozoa. The latter are distributed as fol- lows: There are 9 rhizopods, almost 2,400 ciliates, and over 15,000 flagellates for each one of the Cladocera. There are also about 86,000 plants for each of these Cladocera. Of these plants, 64,000 are diatoms, 14,000 are Schizophycece, 9,000 Chlorophycece, while but 8 are desmids. The great abundance of diatoms, of green and blue-green algae, and of chloryphyll-bearing flagellates affords, it would seem, an abundant food supply for the zooplankton. If of the Mastigophora the colorless flagellates only be retained in the zooplankton, and the remainder—which are 'predominantly synthetic forms—be included with the phytoplankton, we find the latter outnumbering the analytic organisms (zooplankton) 18 to 1. Quantitative values in the matter of food relationships are not readily determined except by a combination of the chemical and experimental method. These results by the statistical method express, with more or less error, the equilibrium of the biological components in terms of the individual organisms. Discussion of the Statistical Data of the Species composing THE Plankton of the Illinois River in 1894-1899. In the following pages the organisms occurring in the plankton of the Illinois River will be recorded, and from the statistical data accumulated by the enumeration method, facts pertaining to their relative abundance, seasonal distribution, and periods of max- imum occurrence will be cited. The average number per cubic meter for the year 1898 will be given, based upon the averages of 52 collections distributed regularly throughout the year (Part I., Table III.). This year is chosen because of the regularity of the times of collection and the absence of any considerable irregularity in the hydrograph. Statements concerning seasonal distribution, etc., are based upon the records for all the years—1894-1899. All figures pertaining to species or groups marked w^ith an asterisk, and starred figures elsewhere, are based upon filter-paper catches; all others, upon those of the silk net. Temperatures are in Fahren- heit, and are of surface waters at time of collection. The margin of error in statistical work of this sort is confessedl}'' large. The complex character of the data with which I am deal- ing, and especially the extreme range in numbers, have made it necessary that I should adopt some consistent method of treating the computations. I have therefore chosen to carry out the num- bers to units, as the most feasible method of avoiding confusion in the handling of the data. The use of round numbers would have been just as accurate. Computation to units is therefore to be understood as a matter of convenience, and not as an effort to exhibit a false and unattainable accuracy. cryptogamia. bacteriace^.* Records were kept of the masses of the larger members of this group which occurred in our plankton catches. They were princi- pally the dichotomously branched brownish fragments of Creno- thrix, filaments of Beggiatoa. and colonies of Micrococcus. The average number recorded for this year was 57,142,822, and they occur throughout the year in every collection, rarely falling below 18 19 10,000,000 per m.^ and reach their maximum development (over 600,000,000) in winter months (December to February), especially during low water and more stable conditions, as in January, Feb- ruary, and December, 1898 (Pt. I., PI. XII.). At such times the temperature is at or near 32°. With flood conditions and rise in temperature the numbers fall below 100,000,000, running from 10,000,000 to 50,000,000 during most of the summer. The decline is due in part to the dilution by flood waters, and largely to the retreat up the stream of the crest of the wave of bacterial activity caused by the Peoria pulse of sewage. As noted in the discussion of the chemical conditions, in Part I., this wave hes considerabh' above Havana during the warmer months. Summer floods, as in June and September, 1897, are wont to wash into the river large C|uantities of these organisms, bringing the numbers up to 300,- 000,000 at times. The figures above cited give but a feeble repre- sentation, of the real conditions in the river during this period of maximum. Many of these organisms become attached to objects along shore, and accumulate in great quantity in quieter waters along the channel. They form a serious menace to the fishing industry, since they accumulate in a day or two upon the fyke-nets in quantity so great that their weight and resistance to the current are sufficient to break down the nets. Their effect upon the consti- tution of the plankton is seen in the marked increase in certain ciliates which accompanies the maximum of these organisms. SCHIZOPHYCE.^. Nine forms were recorded, though a number of others which occurred but rarely in the plankton remained unidentified. The average number (combined silk and filter-paper records, but omit- ting the former when the latter are available) is 85,909,985 per m.^ This group contributes to the plankton throughout the year, and though numerically abundant is quantitatively less important, owing to the small size of its most abundant member. Microcystis. This species and Oscillatoria constitute quantitatively the greater part of the blue-green algae of the plankton. In contrast with the plankton of Lake Michigan, there is a noticeable decrease in the proportion of Anabcsna and Clathrocystis. Rivularia, Gloiotrichia, and Aphanizonienon fios-aqucc, often reported in fresh-water plank- (3) 20 ton, were not found in our fluviatile environment. This group contributes to the water-bloom, contains a number of adventitious planktonts, and is one of the primal sources of the food supply. In our waters it seems to be quantitatively much less important than either the Chlorophycecs, the BacillariacecE, or the synthetic Mastigophora. DISCUSSION OF SPECIES OF SCHIZOPHYCE^ . Anabccna spiroides Klebahn.*—Average number, 637,692 (silk 15,431). In the water-bloom from the last of June till the end of October. Not noted in 1898, but not infrequent in 1897—a low^- water year. Temperature range, 60°-89°. Data insufficient to determine maximum. Largest number recorded, 7,200,000, June 28. Clatkrocystis csruginosa (Kiitz.) Henfr.—Average number of colonies or masses, 83. More abundant in the previous low-water year. From May till the end of Novemiber in the water-bloom. Predominantly a midsummer species. Maximum in August and September (108,000) . Confined principally to the low water of mid- summer, appearing when the water reaches a temperature of 70°, and reaching its maximum development in temperatures above this point, declining at once to small numbers (less than 1,000) when the temperature falls below 60°, but lingering till the water approaches the freezing point late in November. Merismopedia glauca (Ehrbg.) Nag.—Average number of col- onies, 93. In 1897, 889,412.* "In the water-bloom. Recorded from July till the end of October, and also singly in January and February. It was more abundant in 1897 than in 1898, and the maximum number (15,840,000*) appeared on August 31. Microcystis ichthyoblabe Kiitz.*—Average number, 83,059,615. Recorded in all collections throughout the year, except in some flood waters of February and March, when the silt probably ob- scures it. Minimum numbers (less than 50,000,000) prevail during cold months, November to April, when the temperature ranges from 32° to 50°. A well-sustained pulse exceeding 200,000,000 appears with the volumetric plankton maximum of April-May (Pt. I., PI. XII.) and declines to the previous minimum with the falling off in the plankton. The maximum pulse appears later, in August and September in 1898, in September and October in 1897, 21 averaging about 200,000,000, and reaching 1,697,000,000 August 9, 1898. The temperatures diiring these pulses are above 60°, and the period of the maximum comes toward the close of that of max- imum summer temperatures, and sometimes in the autumn decline (Pt. I.. PL XI. and XII.), when low and often stable river-levels usually prevail. A ^-ernal and an early autumnal pulse are thus both present in the distribution of this species. It is not improb- able that other species than the one named have been included in the enumeration along with it on account of the small size and lack of striking characteristics. There are suggestions of recurrent pulses at intervals of 2-6 weeks in the records (Table I.). Oscillatoria spp.—Average number, 15,431 (filter-paper, 637,692). The probable inclusion of several species in the sums under this heading may account in part for the irregularity of the seasonal curve. Oscillatoria has appeared in every month of the year, though the occurrences were most frequent in the period from July till the first of October. The numbers are exceedingly irregular and variable, and the pulses of numbers seem to attend the initial stage of floods following stable conditions. Thus, while these organisms occurred but singly or sparingly in the plankton during the autumn of 1897, they rose to 277,200 with the flood of January 11. 1898, doubtless torn loose by the current from the bottom — their normal habitat. They are thus usually adventitious addi- tions to the plankton. Their frequent irruption into the plankton during midsummer and early autumn, and to some extent at other times, is due in part to the evolution of marsh gas in the detritus on the bottom. This breaks up the mats of Oscillatoria which coat the bottom and distributes them through the upper levels, where they remain in suspension for some time. This phenomenon is more prevalent in the marshy baclcwaters than it is in the river. Flood invasion in midsummer into the backwaters, such as Quiver Lake, is wont to cause there stagnation and great increase in Oscil- latoria, which to some extent enters the river with the run-off of the flood. ^Movements in the water and the evolution of marsh gas are thus principally responsible for the presence of Oscillatoria in the plankton. It still remains possible that its flotation during periods of optimum conditions of gro^vth may be due to internal physio- logical conditions which lower the specific gravity of the organism. Its great abundance at times in upper levels in the backwaters sug- 22 gests the action of this factor, and if this be true, it becomes a tem- porary rather than an adventitious planktont. Temperatures seem to bear Httle relation to the occurrence of Oscillatoria in the plankton. Tetrapedia emarginata Schrod.*—Average number, 242,308. From the first of August till the end of October in numbers from 1,000,000 to 3,500,000 per m.^ appearing later and in larger num- bers in October in 1897 than in 1898. At temperatures above 65°. Tetrapedia gothica Reinsch, GUeocapsa polydermatica Klitz., and Glceocapsa sp. were recorded once or twice in the midsummer plank- ton in relatively small numbers. CHLOROPHYCE^. (Plates I. and II.) Average number, 53,175,105, including, without duplication, species from both silk and filter-paper collections. In 1897 this was very much greater (139,739,850), owing to the prolonged low w^ater and higher temperatures of the late autumn. Although abundant, these organisms are outnumbered by the diatoms six to one, and by the synthetic Mastigophora by about two to one. The ChlorophycecB of the plankton, with few exceptions, are minute, and generally escape through the silk net. Pediastrum and colonies of Botryococcus are about the only species of which the usual method of plankton collection in our waters affords a fair representation. The ChlorophycecE appear in every collection examined through- out all the years of our operations, with the exception of eight in midwinter floods in 1895 and 1896. As a group they are adapted to the whole range of temperatures, and exhibit in 1897, on April 28, a well-defined vernal pulse of 367,200,000, and a series of autumnal pulses culminating September 21 at 216,000,000, October 19 at 367,200,200, and November 23 at 52,000,000. In this year the midsummer pulses are of minor importance in comparison with those of spring and autumn. In 1898 the vernal pulse is also well defined, culminating May 3 at 212,406,400, and it is followed by a series of four midsummer pulses of considerable magnitude, which culminate June 14 at 46,000,000, July 19 at 277,000,000, August 9 at 370,000,000, and August 30 at 189,000,000. The autumnal pulse appears September 27, attaining 70,526,400. The summer and autumn hydrographs of this year are much more disturbed than in 23 the previous year (of. PI. XI. and XII., Pt. I.), especially at the time of the autumnal pulse. This may account for the contrast in the two years. The ChlorophycecB as a whole exhibit (PI. I. and II. and Table I.) the tendenc}'' to form a seasonal curve of recurrent pulses at approximately monthly intervals (three to six weeks), which gen- erally coincide with those of other chlorophyll-bearing organisms. Thirty-three forms of ChlorophycecB were recorded, and closer inspection of the collections will undoubtedly yield a considerable additional number either of closely related, and therefore included, species, or of those which occur but occasionally or in small numbers in the plankton. Numerically the leading species in the order of their importance are Scenedesmiis qiiadncauda, Cnicigenia rectangularis , Actmastrum hantzschii, Raphidiiim polymorphum, Scenedesmus genuinus, S. ohli- qims, Richteriella botryoides, Ophiocytium capitatum, Oocystis naegelii, Ccelastrum cambricum, Oocystis solitaria, and Schroederia setigera. With the exception of Botryococciis bra unit and the species of Pedias- tritm, the remaining forms are both quantitatively and numerically of minor importance. The species just named were enumerated only in the silk-net collections, and coenobia rather than individual cells were listed. If allow^ance is made for the loss of small individuals through the silk, and for the increase that would follow if individ- uals rather than coenobia were the basis of representation, Pedi- astriim would occupy a place in the front rank of importance in the ChlorophycecE of the plankton numerically as well as quantitatively. As quantitative factors in the ecology of the plankton, Pediastrum, Scenedesmus, Ccelastrum, and Botryococcus tSike precedence over the smaller, though more numerous, forms, such as Raphidium and Cnicigenia. The group is thus well represented in our plankton both in species and individuals. The leading planktonts of the group reported in European and other waters in lakes and rivers are here represented almost without exception by identical or closely related species. Botryococcus alone seems to be less abundant than in lakes—at least, according to my own observations, it is much more abundant in the summer plankton of Lake Michigan than in that of the Illinois River. The maximum numbers of Pediastrum reported b}^ Apstein ('96) for Dobersdorfer See in July, ^^•hen reduced to number per m.^ are frequently equaled or surpassed in our waters. 24 Data for comparisons in the case of the more minute organisms which escape the silk are lacking, since results of supplementary methods have not, up to the present, been published elsewhere. It seems probable, however, that the ChlorophycecB will be found to be somewhat more characteristic of the plankton of rivers than of lakes, and to be more prevalent wherever the shore with its decay- ing vegetation forms a large factor in the environment or where sewage contamination- affords the requisite food for their develop- ment. DISCUSSION OF SPECIES OF CHLOROPHYCE^. Actinastrum hantzschii Lagerh.*—Average number, 199,038 (silk net, 338). From May until the middle of November, with maximum of 21,600,000 on August 30, 1898, and of 122,000,000 on September 21, 1897. There are also indications of a vernal pulse, which on May 25, 1897, attained 90,000,000. The major pulse occurs late in the summer, in August and September, w^hile dimin- ished numbers continue until the first of November. Three single occurrences were noted in January, 1898, following the unusual prevalence of 1897, but aside from these the species occurs in the plankton at temperatures above 45°, and both pulses lie in temper- atures above 65°. As in many other species, a greater development was attained in 1897, in stable low water, than in 1898 in disturbed hydrographic conditions. This species occurs in the water-bloom, is favored by stable conditions, and finds its optimum temperature between 65° and 80°. Botryococcus hraunii Klitz.—Average number of colonies, 75. In previous years it was much more abundant, averaging 3,300 in 1897. It occurs from the first of April well into October, though in 1897 it continued until the middle of December. It may thus appear throughout the whole range of temperatures, 32° to 90°, but as a rule occurs above 60°. There is a suggestion of a minor pulse in June, 1896, but not in other years. The major pulse attains 57,200 on August 15, 1896, and 42,000 on September 14, 1897, and appears, with smaller numbers, in August of preceding years. The species occurred but sparingly in 1898. It is found in the water- bloom, and is more abundant in the backwaters than in the main stream. Ccelastrum cambricum W. Archer.*—Average number of coeno- bia, 640,384 (silk, 477). Occurs from the latter part of March till 25 towards the end of November, but principally from May through October. There are but slight indications of a vernal pulse, which on May 25, 1897, culminates at 3,600,000. The major pulse cul- minates at 10,800,000 on August 9, 1898. In the low water and prolonged high temperatures of 1897 the major pulse continues through September, culminating on the 21st at 32,000,000. The average number in this year was about four times as great as in 1898. The temperature limit is 43°, though occurrences are few and numl:)ers small below 65°. The maximum development appears within the period of maximum heat, and towards its close. It is characteristic of the plankton of late summer and early autumn. Crucigenia rectangularis Nag.*—Average number of colonies, 7,153,846. Recorded in all months but March and April, but spar- ingly from November till May. In 1897 pulses appeared in August, September, and October, attaining 32,400,000, 57,600,000, and 118,800,000, respectively. In 1898 there was but a single pulse — in August, of 158,400,000. It was more abundant in the former year. It is present continuously in large numbers from July to October, though in 1897 the impetus of the unusual development was manifested by the continuance of the species even into Janu- ary-. The optimum temperatures lie above 70°, in the latter part of the period of maximum heat, though the species has been found in the plankton throughout the whole range of temperatures. The abrupt decline in numbers occurs between 65° and 40°. It is char- acteristic of the plankton of late summer and early autumn. Golenkinia radiata Chodat.—Average number of colonies, 519,231. It appears most abundantly during the April-May plank- ton pulse (7,200,000) and again, in increased numbers, at the end of August, thus suggesting a vernal and a late summer maximum. It seems to be most abundant at about 60°, a temperature somewhat below the optimum for the two preceding species. Two occur- rences in December, 1896, and large numbers in August indicate its adaptability to the full range of temperatures. Oocystis naegelii A. Br.*—Average number, 207,692. In 1897, much more abundant (average, 4,243,235). Present in numbers (over 5,000,000) from the end of May till the end of September. In 1897, pulses of 10,800,000, 46,800,000, and 24,750,000 appear in May, July, and September respectively. Both numbers and oc- currences are much less in 1898. The optimum conditions thus lie 26 above 70°, though isolated occurrences in March and December in- dicate its presence throughout the whole range of temperatures. It appears to be a summer planktont without the marked prefer- ence for the close of the period of maximum heat noted in some other Chlorophycecc. Oocystis solitaria Wittr.*—Average number, 121,153. In 1897 much more abundant, averaging 2,170,588. In this year it occurs in numbers above 1,000,000 from the end of July till the end of October, reaching a maximum of 36,000,000 on September 21, 1897. Its optimum conditions occur during the latter part of the period of maximum heat, at temperatures approaching 80°. It disappears at 60°, save for isolated appearances in December, at 2)3°— a fact which suggests its persistence in small numbers thoughout the year. It is characteristic of the plankton of late summer,—that is, of low water, high temperatures, and stable con- ditions. Ophiocytium capitatum Wolle*.—Average number, 1,465,385. More abundant in 1897, averaging 2,858,823. Present from the last of April until the beginning of November. There is some indi- cation of a vernal pulse, which on May 25, 1897, attains 3,600,000, and on April 26, 1898, 10,800,000. The major pulse appears in late summer or early autumn, attaining 57,600,000 on September 21, 1897, and 28,800,000 on August 9, 1898. The two pulses are separated by an interval in which occurrences are less frequent and numbers smaller. This planktont thus exhibits the tendency towards seasonal maxima near the average temperature. The greater development in 1897 is followed by a prolongation of the occurrences into November. The optimum temperature appears to be about 60° or above, the vernal pulse appearing at that tem- perature, and the major one at 71°. No records occur below 46°. Pediastrum boryanuni (Turp.) Menegh.—Average number, 4,510. This alga was found in every month of the year, though not in every collection examined. The numbers present fluctuate greatly and are usually much less than those of P. pertiismn, with which it is associated, and with which it fluctuates, often with remarkable coincidence. I have included under this head those individuals in which the coenobium is a plate with no intercellular spaces or only insignificant ones. Individuals are not lacking which serve to connect this species with P. pertusum, and, indeed, with others 27 which have been described in this genus. This genus includes the most abundant of the larger algae in the plankton of fresh waters, and it affords an attractive field for the study of variation by statis- tical methods and for the determination by the experimental method of the eft'ect of environmental changes upon structure. The two groups of individuals included here under P. boryanum and P. pertusum give typical curves of seasonal distribution which are so similar that their combination in a single series would not greatly modify the resultant seasonal curve. In the sum total of all collections P. boryanum (1,034,000) includes about one tenth of the number referred to P. pertusum (10,830,117). A few scattering individuals, generally less than 1,000 per m^, appear at irregular intervals during the colder months, from the first of December until the end of March. The number increases as the tempefature rises, and the species appears in all collections un- til November, when it again becomes irregular in its occurrence in the plankton. The fluctuations in numbers during this period are \-ery marked, the pulses of frequency being set oft' by intervals in which the numbers are small. A slight pulse of 2,120 appears on November 17, 1894. In 1895 the vernal pulse attains the very unusual number of 572,824 in the unusually low water of that year, and the autumnal pulse of September 5 is but 10,600, and is followed by a secondary one on November 27 of 4,081, perhaps as a result of the stable conditions and the abnormally high temperatures (above 45°) which then prevailed (Pt. I., PI. IX). In 1896 the vernal pulse culminates i\Iay 18 at 31,164, while the autumnal pulse is scarcely visible and the numbers throughout the summer are small, as a result, it may be, of the repeated floods of that year (Pt. I., PI. X). In 1897, with few vernal data, the vernal pulse does not appear, though a rise to 8,000 occurs on July 21. The major autumnal pulse culminates on September 14 at 14,400, and another one on October 12 at 6,000, attending the late autumn of that year. In 1898 there are vernal pulses—on May 10 of 6,400 and on June 14 of 32,000. The autumnal pulse on September 27 reaches the considerable number of 65,600. In the winter of 1898-99 Pediastrmn was seemingly absent from the plankton. The pulses are thus somewhat irregular, though there is in this species a suggestion of vernal and autumnal pulses at corresponding 28 temperatures. The optimum conditions seem to lie above 60° and the maximum numbers to occur at or near 70°. Pediastrum pertusum Kiitz.—Average number of coenobia, 44,372. This species appears in the plankton in all months of the year and in almost all of our collections. It is the most abund- ant representative of the ChlorophycecE which is retained by the silk of the plankton net, and is quantitatively an important factor in the ecology of the plankton. The numbers during the colder months, from November to April, when the water is from 32° to 40°, are few, and the sequence of their appearance is fre- quently interrupted. As the temperature rises in April the num- bers increase, and the vernal pulse culminates in a maximum in May or June. There is no indication of the vernal pulse in the scattered collections of 1894. In 1895 the pulse is extreme, reaching 5,264,860 on June 19, in a period of exceptionally low water. In 1896 a pre- liminary vernal pulse culminates May 8 at 23,580 and is followed on June 17 by one of 107,200. In 1897 the few spring collections do not reveal any vernal pulse, while in 1898 a minor one on May 17 reaches 5,600, declines to 600 at the end of the month, and rises again to 56,000 by June 21. These vernal maxima all occur—or at least pass through their period of development—before the water reaches its midsummer temperature of approximately 80°. They develop during the transition from 60° to 80° (Pt. I., PI. IX. to XL). Autumnal pulses during the decline from 80° to 60° appear on Sep- tember 5, 1895, (105,996), on September 30, 1896 (9,200), on Octo- ber 12, 1897 (231,200), and on September 27, 1898 (259,200). In addition to these pulses there are others at irregular intervals during the summer: on July 30, 1894 (154,548), on July 2, 1896 (68,400), on August 15, 1896 (22,000), on July 14 (289,600) and on August 31, 1897 (442,000), and on August 2 (295,200) and 30 (326,400), 1898. The optimum conditions of development thus lie above 60°, and pulses are more frequent in spring and late summer or early autumn near 70°, though they appear somewhat less frequently during the summer in our maximum temperatures near 80°. The cause of these pulses is not conclusively demonstrable from the data at hand, owing in part to the interval between examinations. Daily examinations of the plankton and chemical analyses seem to be desirable for such demonstration. There are indications, how- 29 ever, that certain conditions in the en\-ironment increase the amplitude of the pulses by hastening the rapidity of reproduction of these organisms. Of the fifteen well-defined pulses appearing in' our records of six years, all but three minor ones occur in stable con- ditions, such as pertain to sustained low water. The greater part of these pulses, however, occur in declining floods, when contribu- tions from backwaters are considerable. It may seem ill-advised to refer to the conditions of falling river-levels as "stable"; neverthe- less, they are relatively much more stable than those which attend the in-rush of silt-laden flood-waters, and involve fewer changes in factors of the environment. Save in the matter of the relative con- tributions of backwaters and of sewage dilution they resemble those of sustained low water. These Pediastnim pulses are also re- lated to the nitrate pulses (Pt. I., PI. XLIII.-XLV. and Table X.), but the relation is not uniform. In the majority of instances the pulses of 1896-1898 (during which time chemical anaylses are available) coincide approximately with the crest or decline of in- crease in nitrates. For example, the pulse noted on July 17, 1896, of 107,200 from a previous level of 1,210 on June 1, follows a wave of nitrates progressing for three weeks and culminating on June 9 at 3.25 parts per million—arise from 1.5 (Pt. I., PI. XLIIL). On June 16 the nitrates have fallen again to 2.2, and on the 23d to 2.0, but rise on the 30th to 2.8. Pediastnim responds to these changes by dropping from 107,200 on the 17th to 15,000 on the 27th, and by rising again on July 2 to 68,400. Not all of the fluctuations in the two are concomitant. Some of the most marked pulses of Pcdi- astrum appear at the lowest levels of the nitrates. For example, that of August 30, 1898, of 326,400, follows no nitrate wave, though it coincides with a reduction in nitrates to the minimum of .05. On the other hand, the nitrites had just passed on August 23, an un- usual pulse, to .42, falling again on August 30 to .22 and on Septem- ber 6 to .05 with the passing of the Pediastnim pulse. Pulses of Pediastnim are thus apparently not dependent for their develop- ment upon an abundance of nitrates above the levels shown in the analyses, though a decline in these sources of food or in other forms of nitrogen usually attends these pulses. Pediastnim is but one of many factors among the planktonts, and in the environment, biological and chemical, concerned in these changes, and con- clusive demonstration of its ecological relations must be obtained 30 by the. experimental method. The data here cited are suggestive only; not conclusive. The relation of Pediastrum to the volumetric pulses of the plank- ton is not a constant one, though there is some correspondence in their fluctuations. The extreme maximum (3,264,800) of June 19, 1895, is coincident with a plankton pulse of 30.42 cm.^ but the num- ber of collections is insufficient to show the relative fluctuations of the plankton and Pediastrum at that season. In May and June, 1897, and in October, 1898, the Pediastrum pulses culminate shortly after the volumetric pulses. In July and September, 1897, and in August, 1898, they coincide. Polyedrium trigonum Nag.*—Average number, 432,692. Ap- pears from June through September, disappearing when falling temperatures reach 60°. In 1897 it continues through October with the higher temperatures (averaging 65°) of that year. There are slight indications of a September pulse. Polyedrium trigonum forma minus Reinsch and var. tetragonum (Nag.) Rabh., P. bifurcatum Wille, and P. gracile Reinsch, were also recorded in a few collections during the period of occurrence of P. trigonum. They are all evidently summer planktonts. Raphidimn polymorphum Fresen.*—Average number, 21,450,000. Occurs in every month of ^he year and in a majority of the collec- tions. In 1897 a vernal maximum of 201,600,000 occurs on April 27 and an autumnal one of 28,800,000 on September 21. In 1898 a vernal pulse culminates May 3 at 24,000,000, and thereafter throughout the summer at intervals of three to six weeks there occur five other pulses, the greatest of which culminates July 19 at 75,600,000. A pulse of 90,000,000 on a declining flood in February, 1899, indicates an adaptation on the part of this organism to the whole range of temperatures. A pulse of 25,200,000 December 3, 1896, further illustrates this adaptability. Records in 1897 and 1898, however, suggest that the optimum lies above 60°. It is thus a perennial planktont. Raphidium longissimum B. Schroder.—Appeared sparingly in February, August, October, and December, suggesting that it has also a perennial distribution. Richteriella botryoides (Schmidle) Lemm.*—Average num- ber, 6,399,705 (in 1897). From May to November, with a vernal pulse of 25,200,000 on May 25, and an autumnal one of 100,800,000 31 on September 21. Optimum temperature about 70°, and disappear- ing from our records below 60°. Scenedesmus bijugatus (Turp.) Kiitz.*—Average num]per, 155,769. Sparingly from May till the close of September, with slight traces of vernal and autumnal pulses. Scenedesmus denticulatus Lagerh.*—Average number, 86,538. A few occurrences in late summer and early autumn. Scenedesmus genuiniis Kirchner.*—Average number, 778,846. From May till the first of October, but continued through this month in 1897. Vernal pulse not observed, though the autumnal pulse attains 28,800,000 on September 21 and October 26, 1897. Midsummer pulses appear in 1897 on July 14 (16,200,000), August 17 (14,400,000), and in 1898 on August 9 (19,800,000). Optimum temperatures lie above 60°, though an occurrence in December in- dicates the adaptability of this organism to lower temperatures. Scenedesmus obliquus (Turp.) Kiitz.*—Average number, 1,505,769 (silk, 673). This form appears in our records from the last of April until the middle of November. Traces of vernal and autumnal pulses appear in both 1897 and 1898, with intervening midsummer fluctuations of even greater magnitude. In 1897 the vernal pulse on May 25 reaches 3,600,000; a midsummer one on August 10, 5,400,000; and the autumnal one appears twice, once on September 21 at 28,800,000, and again on October 19 at 25,200,000. In 1898 the vernal pulse appears May 10 at 1,800,000; midsummer ones, on July 19 at 10,800,000, and August 9 at 36,000,000 ; and the autumnal on September 9 at 8,100,000. As in some other organisms, these pulses are separated by intervals of three to six weeks. The optimum temperatures lie above 60°, though development begins before that temperature is reached, and the impetus of the autumnal pulse, or acclimatization to lower temperatures, carries the species beyond this limit into temperatures of 45°. There is a marked absence of pulses below 60°. This seems to be a summer planktont with no marked preference for the lower temperatures of spring and autumn. Scenedesmus qiiadricauda (Turp.) Breb.*—Average number, 9,276,923 (silk, 8,611). In this species, as in the case of others of the genus and of the ChlorophycecB generally, the numbers present in 1897 were much greater than in 1898 (32,492,647,* silk, 5,818). Prolonged low water and concentration of sewage afforded stable 32 conditions and food requisite for such development. This species appears in our collections in every month of the year, though in much smaller numbers and less frequently from November to April—that is, below 50°. Pulses of noticeable magnitude appear only above this temperature, and usually above 60°. Slight traces of vernal and autumnal pulses appear in the col- lections of the silk net in 1894-1896. In the filter-paper collections of 1897-1898 they are well defined. The vernal pulse appears in 1897 on May 25 at 46,800,000, and in 1898 on May 10 at 70,200,000. The autumnal maximum in 1897 is remarkable both for its large numbers and its prolongation, culminating twice—first on Septem- ber 21 at 151,200,000, and again on October 19 at 154,800,000. This remarkable development, combined with the stable conditions and higher temperatures (Pt. I., PI. XI.) of that low-water autumn, is responsible for the continuance of the species in our collections throughout the winter. In 1898 the species declined earlier, in November, and was but sparingly represented in collections of the winter of 1898-1899. As in other species of the genus and other Chlorophycecu, midsummer pulses appear at intervals, often of four weeks, but ranging from three to six. In 1897 these occurred on July 14 at 55,800,000 and on August 31 at 21,600,000. In 1898 they appear on June 28 at 10,800,000, on July 19 at 79,200,000, on August 9 at 39,600,000, and on August 30 at 54,000,000. At inter- vals between the pulses the numbers decrease, and in the regular collections of 1898 the minima between the pulses do not in any case exceed 30 per cent, of the adjacent maxima, and are usually very much less. The distribution of the pulses of this species coincides very closely with that of the other species of the genus, and also with that of other ChlorophycecB. For example, Pediastmm pertusum, the most abundant of the larger algae, has seven of its thirteen pulses on the same dates with those of Scenedesmus quadricauda and three others on adjacent dates, leaving but three which are not practically coincident. The operation of some common and general factor in the environment is suggested by such phenomena. The wide seasonal range of this organism gives it a claim to rank as a perennial planktont, though its quantitative distribution shows clearly that the optimum temperatures for its growth lie above 60°. The largest number recorded in 1897 appears October 19 at a temperature of 65°, and in 1898 on July 19 at 84°. It is 00 thus predominant only during the warmer part of the year; and while autumnal and vernal pulses occur, there is no sustained mid- summer minimum intervening between them. The pulses in Scenedesmiis as a rule follow the \-olumetric pulses as shown in silk- net catches (Pt. I., PI. XI. and XII.). Thus in 1897, on September 14 and 21, the plankton measures 19.8 and 3.0 cm.^ per m.-\ respec- tively, Scenedesmiis qiiadricauda numbering 20,700,000 and 151,- 200,000; and, again, on October 5 and 19 the plankton measures 12.92 and 1.86 cm.^ and this alga numbers 93,600,000 and 154,- 800,000. Its share in the volumetric pulses is thus indirect to a large degree, and is perhaps modified by food relations. Schroederia setigera (Schroder) Lemm.*—Average number, 21,450,000. In 1897, 69,040,912. It appears in all months of the year and in almost every collection. It has well-defined vernal and autumnal pulses separated by the summer period, in which only minor pulses occur. In 1898 midwinter numbers are as high as those of midsummer. Schroederia is thus truly a perennial plank- tont. The vernal pulse appears in 1897 on April 27 at 302,400,000, and in 1898 on May 3 at 150,000,000. The autimmal pulse in 1897 culminates on September 21 at 565,200,000, and is followed by sec- ondary culminations on October 26 at 136,800,000, and on Novem- ber 23 at 203,400,000. In 1898, when hydrographic conditions were less stable, the autumnal pulse reached only 50,400,000,—on September 6. This is followed by minor pulses, declining to a mini- mum in the following February. It disappeared in the collections with the flood waters of March, 1899. The sequence of these sec- ondary pulses follows much the same course as has been described for other species, namely, maxima at intervals of approximately a month (two to six weeks) separated by more or less sharply defined minima. There are twelve such pulses (including the major ones) in 1898 and an interval of seven weeks in March-April in which none occurs. Six pulses appear in the last five months of 1897. The optimum temperatures as indicated by the position of the vernal (60° in both 1897 and 1898, as shown in Table III., Pt. I.) and autumnal (71° in 1897 and 79° in 1898) pulses lie between 60° and 80°. This appearance of the vernal pulse at a lower temperature than the autumnal (usually about 10° lower) is not confined to this species but is a general phenomenon among other Chlorophycece. 34 It is apparently a phenomenon of seasonal acclimatization, by virtue of which the low temperatures of the winter lower the optimum for the vernal pulse, and the high temperatures of the summer raise it for the autumnal pulse. Selenastrum hihraianum Reinsch.*—-Average number 519,235. Recorded only from the beginning of August till the end of Novem- ber, and never in great abundance. Slight evidence of a September pulse. Some other ChlorophycecB have been included in the totals as "unidentified," and isolated occurrences of the following have been noted: Cerasterias longispina (Perty) Reinsch, C. raphidioides Reinsch, Dactylococcus infusionum Nag., Glceocystis gigas (Kiitz.) Lagerh., Staurogenia lauterborni Schmidle, and a few of the Con- fervacecB—which are probably adventitious. These are a species of Conferva, of Prasiola, and of Ulothrix—all of which appear sparingly in spring and autumn planktons, the first-named and the last as minute filaments in the filter-paper collections. A thorough analysis of the unidentified forms would greatly extend the list of species and varieties. BACILLARIACE^. (Plates I. and II.) Average number, 396,192,716, including, without duplication, diatoms from both silk and filter-paper collections. They were almost twice as abundant in the more stable conditions in which the collections of 1897 were made. The BacillariacecB are more abun- dant than any other synthetic group of organisms in our plankton. They exceed (in 1898) the SchizophycecB five to one, the Chloro- phycecE seven to one, the desmids eight thousand to one, and the synthetic Mastigophora by more than four to one. Their numerical preponderance is, with the exception of the synthetic Mastigophora, equaled or exceeded by their relative quantitative significance in the ecology of the plankton. They appear without exception in every collection, and their seasonal distribution in its main features is repeated from year to year. There is a principal vernal pulse in April-May and a hiemal pulse in November-December. Minimum periods separate these pulses and are varied by other pulses, usually of minor importance, at intervals, in 1898, of three to five weeks. The winter minimum 35 is at a lower level than the summer one. In 1894 the interval of collection is too great to follow the seasonal distribution, but there are hints of summer and autumnal pulses. In 1896 there were no May collections, and the largest number, 6,060,665, appears June 19, five minor pulses—on July 18, August 21, September 12, October 11, and November 5—intervening before the hiemalpulseof 3,574,- 028 appears on November 27. Other pulses follow on December 18, January 6, February 4, March 4, and March 17, before the vernal pulse of 1896 culminates at 105,440,858 on April 24. This is fol- lowed by minor pulses on May 18, June 1 1, July 18, August 8, and September 16, and by the hiemal pulse of Decembers of 346,982,- 928*. The vernal pulse of 1897 appears April 27 at 6,207,473,520, but is surpassed by a pulse on July 14—principally of Melosira spinosa—of 11,459,289,600, and minor pulses then follow on August 17, September 29, October 26, and December 7 and 21. The hie- mal pulse of this year is insignificant. In 1898 three minor pulses appear, January 21, February 15, and March 22, and the vernal pulse culminates May 10 at 3,865,257,360. Minor pulses follow on June 14, July 19, August 9, August 30, September 27, October 25, and November 22, and the hiemal pulse culminates December 15 at 436,535,790, followed in 1899 by minor ones on January 10, Febru- ary 14, and March 14. Some of the pulses here indicated are due to the development of single species, as that of Melosira on July 14, 1897. Most of them, however, are composite, including a number of species. This is especially true of the vernal pulse, which in 1898 is due to the com- bined increase in Fragilaria virescens and F. crotonensis, Cydotella, Astertonella, Navicula spp., and Synedra acus. Asterionella culmi- nates early in the vernal pulse and the majority of the others towards its close. Melosira varians is among these, but M. spinosa contributes less to this pulse than it does to later ones. Minor pulses are also composite, as, for example, that of August 9, 1898, which is due to Melosira spinosa, Cydotella, and Navicula. FACTORS CONTROLLING DIATOM PRODUCTION. The fact that many of these pulses represent the combined fluctuations of a number of species leads us to look for some factor * Filter-paper collections included in this and in following years. (4) 36 in the environment common to them all to which these pulses may be attributed. On the following page the seasonal distribution of the total diatoms has been plotted for 1898, along with that of the ni- trates and of the total plankton (volumetric), the thermograph, and the hydrograph. An examination of the changes in nitrates yields no marked evidences of correlation. The vernal pulse of diatoms follows the high nitrates of winter and spring, and the hiemal pulse in December appears after their autumnal rise, and in this particular year develops at the time of an unusual drop in nitrates (Pt. I., PL XLV.). The diatom pulses do not show any constant relation to the movement in nitrates either in amount or direction. Whipple ( '94) has noted the importance of nitrates in the development of diatoms in reservoir waters. The fact that little correlation appears in our waters between the fluctuations of the nitrates and the growth of diatoms may be due to the presence here of nitrates —owing to sewage contamination—far in excess of the demands which the diatoms make, and the limitations placed by other elements in the environment are reached before that of the nitrate food- supply becomes operative. The distribution of these diatom pulses throughout the whole year, even in seasonal extremes, seems to pre- clude the factor of temperature as the immediate cause of the pulses except as it may affect the grow^th of individual species, which is sometimes apparently the case, as is shown in subsequent pages. The vernal pulse is attained each year about May 1, at which time the water passes the temperature of 60°. The average of the recorded surface temperatures of 1898 in the river is about 58°. Surface temperatures, except in winter months, are usually several degrees higher than bottom temperatures (Pt. I., Table III.). Our records are always of diurnal temperatures. The true average tem- perature, owing to colder water at lower levels and to the nocturnal decline, will lie several degrees below 58°—probably about 55°. The greatest development of diatoms thus takes place at a temper- ature a few degrees higher than the average temperature for the year. Owing to the somewhat greater abundance of diatoms dur- ing the warmer months, the average thermal exposure of the plank- ton diatoms will be somewhat higher than the average temperature of the year. There may be some significance in this phenomenon of the occurrence of the optimum temperature for development at 3; oom GQ U -8 2 •- cr X f— Q- 2: CD ^ dco s ei in wo 3S approximately that of the average thermal exposure. The vernal pulse may, in part at least, be the result of a process of natural acclimatization. The fact that a similar development does not recur when this temperature is repassed in the autumnal decline militates, it is true, against the potency of this temperature as a factor in the vernal pulse. This temperature is passed in October (Pt. I., PI. VIII.-XIII.), but October pulses are rarely so pronounced as those of adjacent months. Other factors more potent than tem- perature are operative at that season of the year. As will be seen in the diagram, the most pronounced and pro- longed minimum appears in January, February, and March. In these months but a single record in excess of 100,000,000 per m.^ is found. This—-or at least the first two months of it—is the period of the ice blockade (Pt. I., PI. IX.-XIIL), during which the aeration of the water by the wind is prevented, and the customary equilib- rium in gaseous contents may be disturbed. It is the time when stagnation most threatens disaster to the plankton. The earlier stages of this blockade in December do not seem to be deleterious to the growth of diatoms, since at such times the blockade is less complete, the exclusion of light by the ice less effective, and the accumulation of the products of decay less pronounced. The data at hand do not suffice to elucidate the matter further. The position of the diatom pulses with respect to the movement of the hydrograph is suggestive—though not conclusive—-of a pos- sible correlation between the two phenomena. The double vernal pulse of April-May appears in the declining waters of the major spring flood. The diatom pulse of June 14 is found in the decline of the May-June flood. The pulse of August 9 is caught on the ris- ing waters of a slight flush of the river, and that of August 30 on its decline. That of September 27 appears after a series of slight rises, and those of both October and November attend rising water, but the well-developed pulse of December appears with its decline. There are, counting the double vernal pulse, ten pulses in T 898, from March to January. Of these, seven are found on declining floods, and but three on rising water, and two of these three appear during the slow rise of October-November. Furthermore, the magnitude of the flood is correlated with that of the diatom pulse. The vernal pulses of 3,453,778,080 and 3,865,257,360 attend the major spring flood, culminating April 2 at 18 feet; the pulse next in 39 size, that on June 14 of 1,039,619,680, attends the decline of the flood next in importance—that culminating May 25 at 13.9 feet; while the third pulse, that on December 15 of 436,535,790, attends the decline of the flood culminating November 25 at 8.7 feet. The hydrograph of 1897 (Pt. I., PI. XI.) is unlike that of 1898 (Pt. I., PI. XII.) in the delay of the so-called "June" rise, which culminates July 5 at 7.5 feet. Its decline runs through the month into August. The diatom pulse attending the "June" rise of 1897 appears about a month later than it did with the earlier pulse of 1898, culminating July 14 at 1 1,459,289,600. A delay in the flood is thus attended by a delay in the diatom pulse. In 1897 there is no December rise and no diatom pulse of noticeable magnitude, though in, 1895, in similar absence of the flood, there is a well-defined diatom pulse. In 1896 there is a series of five floods, each involving the early stages of overflow (Pt. I., PI. X.), and on the decline of each occur one or more diatom pulses. It is but natural that the greater number of diatom pulses should fall on declining river-levels, since, as I have previously shown, these periods exceed in duration those of rising floods. They also predominate during the prevalence of seemingly favorable temper- atures, and are characterized by relatively more stable conditions in the environment. There is, however, it seems to me, another and more potent reason w^hy diatom pulses appear at such times. It lies in the overflow of seed-beds in the margins of the permanent backwaters and the run-off of the plankton which develops there with the fall in levels. This is very apparent to one familiar with the locality. During the decline of the flood the channel current is often diverted in minor lateral channels, such, for example, as that (Pt. I., PI. II.) which courses through Thompson's Lake Slough into Thompson's Lake and out again into the river at its southern end by way of "the swale" and the "cut road." A similar current on the eastern bottoms, which enters partially by way of Mud Lake Slough, rejoins the river through Quiver Lake. These lateral currents are joined by the run-off from overflowed bottoms and adjacent marshes and swamps, all of which, as well as the permanent back- waters thus draining into the channel, breed at such times an abun- dant plankton including diatoms. The contributory function of the backwaters to the plankton of the river proper is thus at its maxi- mum during the decline of the flood. 40 As the flood recedes, relict pools on the bottom-lands and along the margins of the permanent backwaters are formed, in which the conditions favoring sporulation or other means of providing for resuscitation are to be found. The emerging bottom-lands thus be- come the seed-bed for starting a new cycle of diatoms whenever flood conditions return. In the river, on the other hand, the conditions for sporulation are not so favorable, and the current tends to carry away such resting stages as may be formed. The observed facts reeardine the distribution of diatoms and the examination of the conditions under which these pulses occur thus alike yield corrob- oration of the view that floods are potent factors in determining the occurrence of diatoms in fluviatile waters, especially where backwaters are extensive. The nature of the action of floods is in some respects similar to that of the overturning of the water which occurs in lakes when the point of maximum density, 3 9 . 2° , is passed in either direction. In lakes of some depth the vertical circulation of so large a volume of w^ater results in a stirring up of the bottom deposits containing the resting stages of diatoms, so that they are brought again into increased light and to better aeration. Whipple ('94) has emphasized the importance of this overturning in starting the growth of diatoms. In our shallow waters this physical phenomenon is of less impor- tance than in the deeper waters of the lake or reservoir. The vol- ume in circulation is smaller, though some compensation for this may exist in the possibility of repeated overturnings with fluctua- tions in temperatures at the critical stage. The existence of cur- rents, the movements of fish, and the roiling effect of strong and long-continued winds upon our shallow backwaters, combined with the fact that much of the seed-bed area of overflow is dry land at the time of the autumnal overturning, all serve to minimize the eft'ect of this overturning in our waters upon the growth of diatoms in the plankton. The spring overturning occurs early in March, and in 1896, 1898, and 1899 a slight pulse not exceeding an increase of 100 per cent, follows the overturning within an interval of a fortnight. The vernal pulse is about two months later than the overturning, and the relation of this to the overturning does not seem to be inti- mate. The autumnal overturning occurs towards the middle or end of November, and in 1895, 1896, and 1898 the hiemal pulse of December follows close upon it, within two, or at most three, weeks. 41 The relation is here more apparent, but the resulting pulse is no larger than those following upon floods during summer, and but Httle larger than the ones which precede it in the autumn. The effect of this overturning upon the plankton of the Illinois River may thus be detected, though it is here of less importance than in lakes and reservoirs since it is overshadowed or replaced by other and more potent factors. The relation of the seasonal distribution of the diatoms to that of the total plankton is not readily unraveled. The latter is the resultant of a most complex series of factors, whose number and relative potency are subject to constant change and readjustment in the unstable environmicnt of the stream. It is the biological expression of the state of tension among these various factors which for the moment exists. Of these factors the diatoms are but one, though an important one, in the food cycle and ecology of the plankton. The v^olumetric determinations in the diagram (p. 37) do not give the true seasonal distribution of the total plankton owing to the escape of an unknown quantity through the meshes of the silk net. They represent more truly that of the animal plank- ton than that of the phytoplankton. A comparison of the seasonal distribution of the diatoms and total plankton may serve, in spite of the errors involved in the volumetric determinations and the disparity of individuals among the diatoms, to throw some light on the effect of the fluctuations of the latter upon the movement in the volume of plankton. A close comparison of the two seasonal curves reveals the fact that the diatom curve is not identical with the vol- umetric curve. It is true that the double vernal (April-May) pulse of diatoms coincides in location with the vernal volumetric pulse. This is also true of the pulses of June 14 and July 19. The crest of the volumetric \'ernal pulse is, how^ever, lodged between the double apices of the diatom curve, and all the stibsequent volumetric pulses from July on lie in depressions of the diatom curve, and vice versa. It is apparent at once on examination of our planktons that the catches of the silk net are from the volumetric standpoint largely, indeed overwhelmingly, of animal origin. These volu- metric pulses are as a rule largely pulses of the zooplankton . It is therefore to be expected that the diatoms would decrease at such times, since they form the food of many Entomostraca and not a few Rotifera. The appearance of the diatom pulses before or after the 42 volumetric (animal) pulse may therefore in a measure present the wavering tendency to establish an equilibrium between these tw^o elements of the plankton. The presence of an abundant animal plankton may therefore be a cause of some of the minimum periods between diatom pulses. Other causes, such as decline of food ele- ments, may also arise, but in our waters the nitrates at least rarely ever reach a le\'el where an unutilized margin capable of support- ing a large diatom population is not still present. Data concerning other food elements are not at hand, but their paucity in water derived from such varied sources and so liberally fertilized by organic wastes seems improbable. There is also the further possi- bility—and, indeed, from the data in hand the probability—of the existence among diatoms of reproductive cycles, interrupted by resting periods. The available data do not, how^ever, throw any light upon the nature of this internal factor or the cause for the running down of the energy of reproduction, and but little upon the operation of environmental factors which stimulate anew the process of reproduction. The seasonal distribution of the diatoms as a whole, and that of individual species also, offer repeated instances of recurrent pulses at intervals approximating four weeks—the lunar month. In 1898 thirteen such pulses can be detected. These often correspond roughly to minor flood intervals, but not always so, for occasionally two pulses occur on the decline of a single flood. Similar appear- ances may be traced in other years, when collections were frequent enough to exhibit minor pulses. They are, however, in all cases quite irregular, and exceptions are frequent. That cosmic factors may indirectly, through immediately environ- ing factors, affect the reproductive phenomena of pelagic organisms has been suggested by the work of Kramer ('97), Mayer ('00), and Friedlander ('01) in the case of the "Palolo" worm, a coral-reef annelid whose seasonal swarming for reproductive purposes occurs at somewhat definite lunar intervals. While the data concerning the seasonal distribution of diatoms in the Illinois River may serve to suggest the operation of an enig- matic cosmic factor, I wish distinctly to state that in my opinion they are wholly inadequate to establish either its presence or its potency. It is much more probable that we have to deal merely with some matter of food relations between the plants and animals 43 of the plankton, and perhaps with the result of increased photo- synthesis in periods of lunar illuminaUon, which tends to establish the limits of the pulses. The number of forms of diatoms noted in our records in the plankton of the Illinois River is thirty-one. This number could be greatly increased by the inclusion of the many adventitious species which flood-waters bring into the plankton and by the addition of rarer limnetic species. Of these thirty-one at least twelve are eulimnetic, while the others are in the main adventitious. There are no species among them peculiar to the potamoplankton, and the dominant forms here are also abundant in the fresh-water plank- ton of our own Great Lakes and of European streams and lakes, barring a few mooted points of specific identity. The limnetic species are fourteen in number, viz. : Asterionella formosa, A. gracillinia, Cyclotella kuetzingiana , Diatoma elongatum va.T.tenue, Fragilaria crotonensis, F. virescens, Melosira granulata var. spinosa, M. varians, Meridion circulare, Rhizosolenia eriensis, Stephanodiscus niagarcB, Synedra acus, S. acus var. delicatissima, and Tahellaria jenestrata. Of these limnetic forms the more impor- tant ones are Asterionella gractllima, Cyclotella, Fragilaria virescens, Melosira gramdata var. spinosa, and Synedra acus and its varieties. The absence or small number of certain limnetic species is notice- able. These are several species of Tahellaria Sind Attheya. On ac- count of the abundance of silt and the transparency of .4 tthcya it may have been overlooked. It has hitherto been reported from waters much nearer the sea, and this coupled with its affinities to marine diatoms may explain its absence in our waters. The remainder of the forms are adventitious, or largely so, and with the exception of the species of Navicula they have little effect upon the ecology or quantity of the potamoplankton. DISCUSSION OF SPECIES OF BACILLARIACE^. Asterionella Formosa Hassall.—Average number of individual cells, 960. Average size of colony, 4.8 cells. Recorded only in November, December, and from February through April, and never in large numbers. The greatest pulse attained at any time cul- minated on March 30, 1896, at 54,540. Aside from an isolated occurrence on June 27, 1896, no individuals were recorded at tem- peratures above 48°, and three fourths of the occurrences are at 44 temperatures below 40°. The data are insufficient to trace the pulses satisfactorily. This species is distinguished with difficulty from A. gracillima, and may include only old, and in our planktons often heavily incrusted, individuals; or it ma}^ be only a low^-tem- perature variety of the species above named, which in the grand total of all our collections outnumbers it ten thousand to one. Asterionella gracillima Heib.—Average number of individual cells, 28,860,160. In 1897 the species was only one third as abun- dant, a contrast which finds its explanation in the fact that the June rise of that year (Pt. I., PI. XL) did not reach the stage of overflow, and a June pulse is absent in the collections of that year. The seasonal distribution of this organism is one of the best-defined and most striking of all the components of the river plankton. It is peculiar in the fact that it appears in numbers only during spring and the beginning of summer, and in the absence of any autumnal pulse upon the return of the temperatures in which the spring pulse ap- peared. This species was recorded in every month of the year but October, but always in small numbers after July 1. In 1894, collec- tions were not commenced until after the tijne of the spring pulse. In 1895 the spring collections were few, and at intervals so great as to preclude the detection of the full course of the spring pulse. The maximum number in the collections of that year appears April 9 at 1,203,100 and falls to 445,995 on April 29—which is approximately the time of the maximum of subsequent years. This was a year of unusually low water during the spring, and overflow stage was at no -time reached (Pt. I., PI. IX.), which may account for the apparent suppression of the spring pulse. The species does not reappear in the collections of that year until December, but it continues in small numbers (less than 5,000 per m.^) until the end of March, 1896, when there is a rapid increase w4iich culminates April 24 at 26,281,400. It disappears entirelyfrom the records at the end of a fortnight, and save for a single entry in June and two in September it does not again appear in 1896. In 1897 the culmination of the spring pulse occurs April 27 at 324,633,600—three hundred-fold larger than in the previous year. There is a normal March flood (Pt. I., PI. XL), on the declining stages of which this pulse appears. With the close of June the species disappears from the records. The June rise does not reach the stage of overflow, and the scanty records show but this single pulse throughout the year. Beyond a single entry in August and in 45 November the species does not again appear in the records during the year. In 1898 there is an unusual midwinter pulse on January 11 of 146,280, followed by a decline and irregularities due to the ris- ing winter flood (Pt. I., PI. XII.). At the middle of March a rapid increase ensues, culminating April 26 at 891,648,000 on the declin- ing spring flood. A decline to 197.683,200 is found at the close of a week, and it is accelerated by the secondary spring flood, which attains the overflow stage of 15 feet in the closing days of May (Pt. I., PI. XII.). With the decline of this flood in June a second pulse appears, increasing from 15,080 on Ma}^ 26 to 336,194,880 on June 14, and at the end of three weeks the species practically dis- appears from the plankton. A few scattered entries appear during the summer and fall, and a minor pulse of 10,500 appears on Decem- ber 20, followed by a decline in the next month. This species in our waters exhibits a well-deflned vernal pulse towards the end of April at about 60°, but no autumnal pulse appears when this temperature recurs. There is a slight indica- tion of a minor midwinter pulse at the minimum temperatures of the year. This occurrence of a midwinter pulse was noted by Whipple and Jackson ('99) in the reservoirs of the Brooklyn water-works, and in the same paper its seasonal distribution in Fresh Pond. Lake Cochituate, and Wenham Lake, Massachusetts, is given for the years 1890-97, in the majority of which a mid- winter pulse commensurate in magnitude with the vernal pulse is to be found. Autumnal pulses are of infrequent occurrence, the vernal pulse being the most frequent but not constant. In European waters no such long-continued examination of the seasonal distribu- tion of this organism has as yet been reported. Apstein ('96) finds two pulses per year in Ploner See—in May and the last of July ; and two in Dobersdorfer See, one in April and one in October, separated by midsummer and midwinter minima. Lauterborn ('93) finds that this species in the " Altwasser" of the Rhine attains its maximum in June and again increases in October. In the backwaters of the Elbe, Schorler ('00) reports Astenonella as abundant in April, June, July, and October, but refers the organisms to the preceding species. The existence of the vernal pulse only in our waters is thus somewhat unique, and the cause of the phenomenon probably lies in some environmental conditions, perhaps in our peculiar bacterial and sewage contamination of the autumn. Our vernal pulses appear on 46 declining floods about the end of April at about 60°. It can not be temperature which limits the occurrence of the species, for this apparent optimum recurs again in October. This is the period of declining nitrates (Pt. I., PI. XLIII.-XLV.), but they rise again in the autumn, and in our sewage-fed waters they contain even in the midsummer minimum a quantity adequate to support an abundant growth of Asterionella . Whipple and Jackson ('99) have found on analysis that Asterionella to the number of 10,000,000,000 per cubic meter yield but .079 parts per million of organic nitrogen. The nitrates inour waters rarely fall below. 25 parts permillion, which, with theother forms of nitrogen that may be available, would seemt to afford nurture not only for Asterionella but also for competing organisms. These authors have also found that silica to the amount of 1.7 8 andmanganic oxide to .03 per million are contained in Asterionella to the number per cubic meter above quoted. As was shown in Pt. I., p. 234, the silica is present in great excess (26 to 81 parts), and the manganic oxide, though not reported in the analyses of November waters, is present on June 15 to the amount of .07 parts per million—more than double the amount required to support Asterionella to a maximum twelve times as great as any recorded in our plankton collections. This also occurs at a season when Asterionella is usually declining rapidly in numbers. Such chemical data as are available thus afford us no explanation of the limitation of Asterionella in our waters to the vernal pulse alone. Some evidence bearing on a factor which may be operative in producing this phenomenon is to be found in the hydrographic con- ditions attending the vernal pulse. As previously noted, this appears each year with the decline of the spring flood. A repetition of the overflow in 1898 at the end of May brought with it a repetition of the vernal pulse of Asterionella in early June. With the decline of the flood the backwaters make their major contribution to the channel plankton, and it is during this period that Asterionella reaches its maximum and also declines. If the spring flood is sup- pressed, as in 1895 and 1896, the spring pulse of Asterionella is cor- respondingly feeble. The environmental conditions are thus more favorable in the impounded backwaters than in the main stream. Whipple and Jackson ('99) have noted in frustules of this diatom the appearance of structures which they interpret as spores. If these are spores, and if the sedimentation of spore-bearing frustules occurs 47 extensively in the relict pools of the emerging bottom-lands, a seed- bed for re-stocking the waters of overflow is formed with each declin- ing flood, and this seed-bed becomes potent only when floods return. The absence of an autumnal overflow and the minor part that the autumnal overturning plays in our shallow waters when 39.2° is passed, may alike tend to suppress here the autumnal or midwinter pulses which occur elsewhere in deeper water. The occurrence of the vernal pulse of Asterionella in the last days of April brings it into close relation with the major volumetric pulse of the year (Pt. I., PI. IX.-XII.). It is not only an important con- stituent of this spring maximum, but it is one of the most prominent primal sources of food of the Eniomostraca—Bosmina, Daphnia, Cyclops, and Diaptomiis, all of which exhibit an increase in numbers at this period. It shares with Cyclotella the claim to the first place quantitatively among the synthetic organisms upon \^hich the early spring plankton depends for its development. Our records are all based upon the catches of- the silk net, through whose meshes the isolated cells of Asterionella readily escape. Filter- paper catches give niuch higher numbers except during the period of maximum, when the num.bers by the two methods do not materially difler. This seems to be due to tTie fact that isolated cells are rela- tively m.uch more abundant after the maxima than they are be- fore them, and especially at the time of their appearance. These diatoms form arcs, circles, or whorls, of a varying number of cells. During the vernal pulses of 1898 the average number in these clus- ters in the middle of March was three or four, and at the time of the maximum on April 26 it rose to five or six, often reaching sixteen or more. A fortnight after this maximum the average fell to 1.4, rising again with the second pulse, on June 14, to 8.4, and decHning in three weeks, with the fading out of the pulse, to 1.2. Asterionella is frequently infested with great numbers of a minute craspemonad flagellate protozoan which appears in thick-set rows upon the ray-like cells, a single cell sometimes bearing a score of these organisms. This diatom exhibits considerable \'ariation in size and proportions. The longer and more slender cells appear at the times of the maxima. Cocconeis communis Heib.*—Average number, 520,000, but more than three times as abundant in 1897. This diatom occurs some- what irregularly in the filter-paper collections, and has been recorded 48 in every month of the year. It is somewhat more prevalent in spring and autumn, and there are indications of a vernal pulse in May and an autumnal one in September, separated by prolonged midsummer and midwinter minima. Vernal pulses appear in 1897 on June 28 at 14,400,000, and in 1898 on May 17 at 7,200,000. Autum- nal pulses occur in 1896 on September 16 at 2,700,000; in 1897 on September 29 at 10,800,000; and in 1898 on September 13 at 5,400,000. The optimum temperatures lie between 60° and 75°, the autumnal pulse appearing in higher temperatures than the vernal as a rule. This diatom is reported as often epiphytic upon algae, and it may be wholly adventitious in the plankton. There is nothing, however, in the curve of its distribution to corroborate this view. Cydotella knetzingiana Thw.*—Average number 243,659,615, but slightly more abundant in the preceding year. This is one of the smallest as well as one of the most abundant of all the diatoms of the river plankton. It readily escapes through the meshes of the silk net, and plankton collections made by this means give no adequate conception of its prevalence or importance in the ecology of the plankton. It appears in every month in the year and in practically all of our collections, and is thus a perennial planktont. There is a considerable variation in size among the individuals in the plankton, but the greater number lie near the smaller rather than the larger limits. It may be that several species have been combined in the enumeration. The fluctuations in the seasonal distribution of this diatom are considerable, and pulses occur at all seasons of the year. The vernal pulse is, however, preeminent, and is not approached in magnitude by those of any other season of the year. In 1897 this pulse culmi- nates at 5,724,000,000 on April 27, and in 1898 on April 26 at 2,880,- 000,000. Throughout the summer and autumn in both years there is a series of minor pulses at intervals of two to eight weeks. In 1897 an autumnal pulse of 223,200,000 appears on September 29, and though not of greater magnitude than two previous summer pulses, it does surpass anything prior to the pulse of the following spring. In 1898 there are seven pulses during the summer and fall, culmi- nating as follows: on May 10 at 2,668,000,000; on June 28 at 291,- 000,000; on July 19 at 561,600,000; on August 9 at 401,400,000; on August 23 at 122,400,000 ; on September 6 at 1 15,200,000 ; on Septem- 49 ber 27 at 57,600,000; on October 25 at 25,200,000; and in December a pulse ^^ell sustained throughout the month culminates on the 15th at 4 14,000,000. The temperature optimum appears to be about 60°, though its return in the autumn does not induce a development comparable with that of the closing days of April. The midsummer pulses and that of December show that other causes than temperature are operative in regulating the occurrence of this organism. The appearance of the vernal pulse of Cydotella at the time of the volumetric maximum (Pt. I., PL IX.-XII.) in. April-May sug- gests its function as one of the primal sources of food for the animal components of that plankton. The plates are based on collections of the silk net, and Cydotella constitutes an insignificant part of the volumetric total there graphically presented, since it is so small that it escapes readily through the silk. Cymatopleura solea (Breb.) W. Sm.*—Average number, 2,115 (silk, 1,292), but slightly more abundant in 1897. Isolated occur- rences in small numbers appear during the colder months, generallv below 60°, though several individuals appear in summer records. This is apparently an adventitious planktont, whose presence is often due to flood waters. Diatoma elongatum var. tenue Van Heurck.*—Average number, 2,471,923. This is a perennial limnetic diatom occurring in every month of the year and in the majority of our collections. It is but sparingly present during midsummer. There are well-defined vernal pulses in 1897 on May 25 of 50,400,000, and in 1898 on May 3 of 18,000,000. A second large pulse appears on the approach of winter, in 1897, on November 15, culminating at 2,700,000, and in 1898, on November 22, at 9,000,000. In the silk collections of 1895 and 1 896 pulses also appear in the last days of April and in Novem- ber or December. The records thus indicate a decided preference of the species for temperatures below 70° and the possibility of rapid development in midwinter—as in 1895, during a fortnight of mdnimum temperatures (32° + ), culminating at 53,424 (silk) Decem- ber 18. The vernal pulses coincide approximately with the volu- metric maximum, and the December pulse of 1895 attends an unusual winter development of the plankton (Pt. I., PI. IX. and Table III.). 50 Diatoma vulgare Bory occurred sparingly at irregular inter\-als, and is apparently an adventitious species in the plankton. Encyonema prostratitm (Berk.) Ralfs appears a few times during the summer months, and is evidently adventitious, as is also the still rarer Epithemia tiirgida Kiitz. Fragilaria crotonensis (Edw.) Kitton.—Average number of cells, 2.1. This limnetic diatom is much less abundant in our waters than the following species. In 1898 it appeared in February, and increased from 19,200 on April 19, to 14,469,120 on May 10, dis- appearing entirely from the records after May 17. Such meteoric pulses were not detected in previous years, when only scattered entries in April, May, and December were recorded. The number of cells in the filaments is very much less than in F. virescens, aver- aging but 14 to its 108. Its optimum temperature lies about 60°, and its vernal pulse occurs immediately after the volumetric maxi- mum (Pt. I., PL XII.) and upon the same date with that of F. vi- rescens. It seems to be predominantly a vernal planktont in our waters. In German lakes Apstein ('96) finds maxima as late as June-July, but always, it seems, at temperatures below 70°. Fragilaria virescens Ralfs.—Average number, 73.1. Apparently ten times more abundant than in 1897, as a result possibly of the absence of collections during the period of the vernal maximum, in that year. This is a perennial organism, with two well-defined pulses ; a vernal one in April-May and another in November- December. The uniformity with which these pulses appeared in 1895-1898 is very striking when one considers the unstable environ- ment in which the pulses occur. In 1894 the species is not present in numbers in any of the scattered collections of the year. In 1895 the vernal pulse is indicated in the collection of April 29 (2,754,675), after which the species disappears until vSeptember, increasing — with a temporary backset by the December flood (Pt. I., PL IX.) — to a second culmination December 30 at 282,225. After a minimum in January, 1896, the numbers increase, with minor fluctuations, to a vernal maximum of 76,224,000 on April 24, followed by a mini- mum period from May 18 to the following November. The winter pulse again appears in December, culminating on the 3d at 867,048. In 1897 the vernal pulse seems to culminate somewhat later than usual, though the interval of collection is too great to follovv' its full course. The maximum appears on May 25 at 3,549,600, after 51 which the species dwindles away and disappears in August to return early in November. The winter pulse culniinates December 14 at 8.159,250, at a break in the ice blockade. In 1898 the winter mini- mum continues into April, and the vernal pulse appears May 10 at 253,960,000, rising with rocket-Hke suddenness from 390,000 of the previous week, and declining the week following to 4,110,400. The decline to the summer minimum is prolonged into July, and the species does not reappear until October. The winter pulse begins earlier than usual, on November 1, and is well sustained through the month, culminating on the 29th at 2,254,000. The winter mini- mum which follows, does not reach the low levels of that of summer. This species has thus a characteristic distribution, the analysis of which is by no means simple. The contrast between the summer and winter minimum may be due to the low nitrates of the summer and the larger amount in the winter (Pt. I., PI. XLIII.-XLV.), which favor a proportionate development of this diatom, though not every species shows this response. The two minima separate the seasonal occurrences of this species into two periods of growth ; a vernal, from March to June, and a hiemal, from October to Janu- ary, the limits and relative development of each being somewhat \'ariable from year to year. The temperatures of the two periods differ. Both are times of rapid change,—of riseandfall respectively, — and the culminations of the periods of growth lie at widely sepa- rated temperatures. The vernal pulses in 1896 and 1898—in which years collections were frequent enough to locate them with some degree of accuracy—appear at 72° (April 24) and 61° (May 10) respectively, and in every year the vernal pulse appears during a period of rapid change. The hiemal pulse, on the other hand, cul- minates in each year after the winter minimum approaching 32° has been reached, and in two years during the ice blockade. Tempera- ture within these limits seems not to be a determining factor in the pulses of this organism. The nitrates (Pt. I., PI. XLIII.-XLV.) have been uniformly high (above 2 parts per million) whenever the pulses occurred. In 1898 they decHne abruptly (Pt. I., PI. XLV) and remain at a low level throughout December, and in this month, when usually Fragilaria attains its hiemal maximum, we find it dropping to the unusual minimum of 20,000. The pulse which began in November is cut off apparently by this unusual decline in nitrates. Abundance in nitrates is not, however, in itself sufficient (5) 52 to cause a pulse of development of Fragilana, for nitrates are abundant when the diatom declines and is at its miinimum. It does not seem possible to find in the unstable environment of this organism any external factor which shows a causal connection with its periods of growth. Apstein ('96) found that this diatom reached its major pulse in March and April in Dobersdorfer See, and a minor one in November. The cells of this diatom form long twisted bands, visible to the unaided eye. They reach a much greater length in this species than in the preceding one, and are longest during the height of the growing period, decreasing rapidly in length as it declines. The average number of cells in a ribbon at the time of the maximum lies between 150 and 200, and at other times is usually below 100 and often below 25. The vernal pulse of this species coincides with that of F. croto- nensis, and appears either with or just after the volumetric pulse. The December pulses may in part serve as primal food sources for the fairly constant minor volumetric pulse of December. Gomphonema constrictum Ehrbg.*—Average number, 501,923. This species appears irregularly, w4th a predominance of occur- rences in May and November, and is apparently adventitious. Melosira granulata (Ehrbg.) Ralfs var. spinosa Schroder. —Average number of cells, 1,181,125 (filter-paper, 34,762,365). In 1897 it was more than five times as abundant. In the filter-paper collections as a whole it is about fifty times as abun- dant as in those of the silk net. A much greater proportion of single cells and short filaments occurs in the latter collections, since the longer filaments are the more readily retained by the silk. In the discussion which follows, the data from the silk collections will be used, since they cover the whole period. The data from the filter-paper collections indicate very nearly the same seasonal routine, and the differences between the results by the two methods lie in the proportions of the numbers rather than in the direction of movement in the fluctuations. The pictures of the seasonal changes in occurrence of the diatom given by the two methods are essentially alike aside from greater irregularity during minimum periods, resulting from the larger margin of error in the filter-paper method as I used it. 53 This Melosira is a perennial planktont in that it occurs in every month of the year in the river. Its appearances from December to ^larch are, however, irregular, and its numbers small. Its large pulses—above 1,000,000—all lie between May 15 and October 1. with the single exception of the pulse of April 24, 1896, culminating at 2,056,400, in temperatures of 72°, occurring fully a fortnight earlier than usual. The major pulse seems normally to occur in June; at least in 1896 and 1898, when collections were frequent at this season of the year, such pulses appear on the 11th at 12,940,000 and on the 21st at 32,114,880. A June pulse also appears in 1895. September pulses appear in 1895, on the 12th, at 2,254,182, and in 1898, on the 27th, at 5,499,840. There is, however, 'no well-defined \'emal and autumnal gro^\i;h period, since large pulses occur through- out the whole summer. The greatest pulse on record (111,456,000) is on July 21, 1897, and in 1898 there are three minor pulses between those of June and September. Including the major pulses, there are in 1895 five, in 1896 six, in 1897 five, and in 1898 eight, pulses at intervals of two to six weeks between May and October, the ones at either end of the season being often but slightly developed, the remainder usually running from 1,000,000 to 5,000,000. This species is predominanth^ a summer planktont, and its optimum temperature lies above 70°, the greatest number recorded appearing at 81°. This is one of the most abundant diatoms of the potamoplankton, and in our waters it attains its greatest de- velopment during the season of the minimum occurrence of nitrates, in whose utilization it is quantitatively an important agent. It fills the gap between the vernal and autumnal or hiemal appear- ances of Asterionella and Fragilaria, thus providing a continuous source of food for the zooplankton with which it is associated. It is, by virtue of its numbers, its size, and its seasonal distribution, quantitatively and ecologically the most important of all the diatoms of the plankton of the Illinois River. The only factor in the environment to which the limitation of the rapid growth of this species to the May-October period can be re- ferred is temperature. There are but three instances in the records of Melosira exceeding 100,000 per m.^ at temperatures below 60°, and one of these is but a few days prior to the attainment of that temper- ature. It cannot be food which deters its development below this point, since the nitrates at least are then most abundant (Pt. I., PI. 54 XLIII.-L.). Other diatoms, as in the hiemal pulse of Fragilaria, develop in numbers at temperatures approaching 32°, but not M. granulata var. spinosa. Whipple ('94) concludes from the records of examinations of potable waters in Massachusetts that temperature has possibly a slight influence on the growth of diatoms, but that it is of so Httle importance that it does not affect their seasonal distrilm- tion ; and, on the other hand, that a sufficient supply of nitrates is one of the most important conditions for their growth. The seasonal dis- tribution of Melosira was not separately discussed in his paper though included in his general statements. In our waters the data at hand seem to show conclusively that abundance of nitrates is of no avail in the case of Melosira when the temperature falls below 60°. There are times, therefore, in the case of this, our most important diatom, when temperature is more potent than food as a factor con- trolling its growth. Melosira does not appear in its maximum pulses at the time of the major volumetric pulse of the total plankton of April-May, nor do its fluctuations seem to bring about directly any considerable changes in the volumiC of the plankton. For example, the extreme pulse of 111 ,456,000 on July 21, 1897, occurs at the time of a sudden drop in the amount of plankton (Pt. I., PI. XL). The amount of plankton on July 14, 21, and 30 is 8.16, 0.92, and 1.05 cm.^ per m.^ and the corresponding numbers, oi Melosira are 66,528,000, 111,456,- 000, and 13,176,000. The diatoms here discussed are predominantly of the type designated as var. spinosa, marked by the spinous prolongations from the valves at the ends of the filaments. The cells of the forms in our plankton are proportionately much longer, as a rule, than those figured by Schroder ('97), usually attaining one and a half to two times the length without proportional increase in diameter. Not infrequently in the height of the growing season much elongated and curved cells and filaments are to be found. In one instance an unusual number of filaments approaching M. varians in form though still of the spinous type were found. It is not improbable that several so-called species of Melosira have been included with this variable species in the enumeration. Melosira is the bearer of numerous passive planktonts, the most abundant of which is Bicosoeca lacustris Clk. Associated with this, and often on the same filament, is the elegant little craspemonad 55 Salpingccca hrunnca Stokes. Cells to which se\-eral of these flagel- lates are attached very frequently exhibit a breaking up of the cell contents into eight brownish masses, often of spore-like form, and it is not an uncommon thing to find such parasitized filaments with several empty cells. The eggs of the rotifer Diurella tigris are fre- quently found attached to the filaments of this diatom. The num- ber of cells in the filaments in the silk collections averages 6.4 in 1897, and 7 in 1898, while in the filter-paper collections it averages 3.5 in both years. The numbers per filament range from 1 to 40, and the filaments are wont to be somewhat longer during rapid growth than in periods of decline or minimum. Melosira varians Ag.—Average number, 148,626 (filter-paper 3,455, 5v38). The discussion is based upon silk catches. The species was about equally- abundant in 1897 but much less so in previous 3^ears. This is a perennial species, reported in every month of the year and in most of the collections. It exhibits two well-defined pulses, a vernal one in April-May and an autumnal one in September- October. The reduction in the minimum intervals varies from sea- son to season and from year to year. It was most pronounced, al- most to suppression, in July and August in 1894, 1895, and 1896, and in December-February in 1896-97 and 1898-99. In other seasons the minimum falls to 1,000 to 15,000. The vernal pulse (146,916) appears in 1895 on April 29, in 1896 (229,235) on May 18, in 1897 (2,419,200) on May 25, and in 1898 (3,164,160) on May 5. The autumnal pulse (150,720) is found in 1895 on October 30; in 1896, on September 16 at 378,900; in 1897 there are two pulses, one on August 30 at 738,000, and the other on No- vember 15 at 458,800; and in 1898 one, on October 18 at 348,000. The autumnal pulses are thus much smaller than the vernal ones and exhibit a greater range in the time of their appearance. As in the case of many other organisms this diatom also exhibits the phenomenon of recurrent minor pulses at intervals of a few weeks. They range in height from 25,000 to almost 1,000,000, and are largest when found in the proximity of the major pulses. The records are not frequent enough to trace them in all seasons. They appear in January in 1896, 1898, and 1899; in February in 1898; twice in March in 1896; in April in 1896; twice in June in 1897 and again in 1 898 ; in July in 1897 and 1898 ; in August in 1897 and 1 898 ; 56 in September in 1898; in November in 1896, 1897, and 1898; and in December in 1894. The optimum temperatures, omitting the pulse of August 30, 1897, at 80°, all lie below 72°, averaging 65° for the vernal pulse and 62° for the autumnal. But three pulses in all, exceeding 100,000, lie at temperatures above 70°, and but three below 50°. In the case of this species likewise temperatures seem to be potent factors in limiting its seasonal occurrence. The fluctuations in nitrates do not seem to bear any constant relation to its develop- ment. The midsummer minimum of the diatom may appear, as in 1896, during an abundance of nitrates (0.5 to 3.0 parts per mil- lion—Pt. I., PI. XLIII.) unusual for the season. On the other hand, a minimum of nitrates (.1 to .35) in August and December, 1898, coincides with a suppression of this species in the plankton. Thus in the presence of food, temperature seems to be a determining factor in the seasonal distribution of this organism. Whipple ('94) expresses the opinion that the growth of diatoms occurs at those seasons of the year when the water is in vertical circulation; that is, when it passes 39.2°. In our waters this generally occurs early in March and late in November. In this species the only pulses which it seems might exhibit the effect of this phenomenon are those of December and March, and neither of them are in any way constant or prominent. Neither of the major pulses, vernal nor autumnal, can be attributed to it. The latter pulse occurs prior to the autumnal overturning of the water. The vernal pulse usualty follows the spring volumetric maxi- mum, and the autumnal one generally appears during a volumetric minimum. No immediate quantitative effect of this species upon the plankton is apparent. In European waters this is a common planktont, and Apstein ('96) reports vernal maxima in March, April, and May, and an autumnal one of munor value in November. The number of cells in the fdaments varies from one to sixty, and in filter-paper collections averages four, while in the silk catches it varies from seven to fifteen from year to year. The fila- ments average somewhat longer during the periods of maximum growth, reaching twelve to twenty-five. This species also occasion- ally bears the flagellates found upon M. granulata var. spinosa, but not in such abundance. It is quantitatively much less important 57 in our plankton than that species, though this does not seem to be the case in some European waters. Meridion circulare Ag. has appeared but four times in winter planktons, from December to March, and seems to be adventi- tious. Navicula iridis Ehrbg.*—Average number, 297,307. Appears at irregular intervals, often with flood waters and in the colder months. It seems to be adventitious. Navicula spp.*—Average number, 8,569,038. About twice as abundant in 1897. Under this head I have included a number of species of Navicula, and, possibly, even species of genera resembling Navicula. The individuals are all of small size, and are principally of the type of the smaller forms of ^V. hrehissonii Kiitz. and N . gracilis Ehrbg. They are quite abundant in collections from Quiver Creek and Spoon River. Their greater abundance in 1898 as compared with 1897 may be caused by the greater movement in river levels in the former year (85.6 ft.) as compared with that of the latter (55.5 ft.). This feature of the distribution of these forms suggests that, they are adventitious in the plankton. This view is further supported by the fact that some, though not all, of their apparent pulses appear with flood waters; for example, the pulse of 64,000,000 on May 17, 1898. There are indications, independent of floods, of pulses in April-May and November-December, which may, how- ever, be simply reflections of pulses in the normal habitat of these diatoms—the shores and bottom of the river and its tributaries. They are represented in the plankton at all seasons, and the diver- gence in numbers is at no time so marked as it is in typical plank- ton diatoms, such as Asterionella. Nitzschia amphioxys (Ehrbg.) Kiitz. appeared several times in ^\inter collections, and A^. sigmoidea (Nitzsch) W. Sm. is adventi- tious in small numbers in flood waters. Several species of Plen- rosigma appear at irregular intervals throughout the year in both flood waters and stable conditions and are apparently adventitious, appearing in relativel}^ small numbers. Rhizosolenia eriensis H. L. Smith was noted on a few occasions in winter planktons. Its exceeding transparency and the abun- dance of silt and debris at the times of its occurrence so obscure it that it may have escaped detection in many instances. 58 Stephanodiscus niagarcB Ehrbg., a common planktont in the waters of the Great Lakes, appeared but once, in May, in our plank- ton, though the river had for years received, by way of the Chicago River, constant access of water from Lake Michigan. The turbid, sewage-laden, and warmer waters of the Illinois are evidently not favorable for its growth. Surirella ovalis Kiitz. var. minuta (Breb.) Kirchner.*—Average number, 761,538. Present sparingly throughout the year, but principally during summer months. Vernal pulse in May. Surirella spiralis Kiltz.—-Average number, 1,612. Less abundant in the more stable conditions of 1897. This species is most abun- dant in Quiver Creek and Spoon River. Its fluctuations are slight, irregular, and often appear with flood waters, all of which phenom- ena indicate its adventitious character in the river plankton. Synedra acus Kiitz.*—Average number, 36,558,462 (silk, 308,- 330). This species is a perennial planktont, appearing, for example, in 1898 in every collection. It has a highly developed and shifting vernal pulse, and an inconstant and but slightly developed autumnal or hiemal pulse. The vernal pulse appears in 1895 on April 9 at 209,880; in 1896 on April 24 at 366,828; in 1897 on May 25 at 2,620,800 (82,800,000*); and in 1898 on May 10 at 9,043,200 (813,600,000*). The second pulse appears in 1895 on November 14 at 99,360; in 1896 on December 3 at 44,464; in 1897 no pulse occurs; in 1898 it occurs on November 8 at 19,000. As in some other diatoms, there are minor pulses throughout the year, though in this case they are all feebly developed, exceeding 100,000 (silk) in but a single instance. The minor pulses of midwinter often exceed in prominence those of midsummer. The meteoric char- acter of the vernal pulse is very pronounced in this species both in the suddenness of its appearance and its disappearance and in the height which it attains. The variety delicatissima W. Sm. is included here with the type acus. During the autumn of 1898 a separate record was kept of the two, with the result that the variety appears to include about four fifths of the individuals at that season. The two are not readily separated. The colorless form recently described by Pro- wazek ('00) as 5. hyalina is also included, and it is not uncommon when 5. acus is abundant. Colorless forms of other diatoms of the plankton, as Asterionella, Melosira, and Fragilaria, also occur, but 59 it would seem from the intergradation with the normal condition that it is a phenomenon of physiological import rather than of specific significance. It would seem desirable that experimental breeding of diatoms should be employed as a test before specific diagnoses utilize this character. Synedra capitata Ehrbg. is occasionally adventitious in the plank- ton in spring months. Synedra ulna (Nitzsch) Ehrbg.*—Average number, 302,308 (silk, 34,510). This appears somewhat irregularly in the plankton, with a vernal pulse on May 17 of 5,400,000 and an autumnal one November 15 of 1,800,000. It is abundant on the ooze of exposed springy shores after rapid decline of the river, and is probably adventitious in the plankton to some extent from this region. Tabellaria fenestrata Klitz., which is exceedingly abundant in the plankton of European lakes and in our own Great Lakes, was found but a single time in the waters of the Illinois. It can hardly be lack of food elements which prevents its development, and there are times when favorable thermal conditions would seem to be offered in spring and autumn, when the river temperatures do not exceed the summer temperatures of our Great Lakes. It may be that the chemical conditions attending sewage contamination exert a dele- terious influence upon this species and others of the genus, such as T. flocadosa, which abound in purer lake waters. CONJUGATE. This group of algae is represented in the plankton only by a few desmids, which neither in number, or quantity, play any important part in the ecology of the plankton. The filamentous algas are abundantly represented in spring in the backwaters of the Illinois River, where they form extensive littoral fringes of "blanket moss," which load down the emerging littoral flora. This fringe is fre- quently stranded by the retreat of flood waters. In some localities, as in Phelps Lake, it plays a very important part in the food cycle, since by its decay, as temperatures approach the summer maximum, it contributes immediately its store of organic nitrogen to the su])- port of the small algfe and flagellates which develop in great num- bers on those waters at that season. Some species of Spirogyra and Zygnema have a habit of breaking up into short filaments, and 60 in this condition they have often been taken in some quantity in the plankton of the river, but they are so plainly adventitious and irregular that no notice has been taken of them in our enumeration work, and when possible they have been removed before measure- ment or deducted by estimation from the volumetric records. The desmids are few both in species and individuals. Seven species have been recognized, of which but four are of general occurrence in the plankton. These are three species of Closterium and Staiirastrum gracile. The latter and Cosmocladium saxonicum are the only eulimnetic organisms among them. The center of dis- tribution of the other species is the shore and bottom. The stom- achs of fish such as the CatostomidcB, the carp, and Dorosoma cepe- dianimi, which often feed upon the bottom ooze or slime about aquatic plants, usually contain many desmids, including the species here noted. Other species also are occasionally adventitious in the plankton, and the list might be considerably extended, though the absence of extensive peat bogs in the drainage basin of the river reduces the desmids to a position of much less importance than that which they occupy in more northerly waters. As a group they exhibit a well-defined seasonal distribution, with a vernal pulse at about the time of the volumetric maximum in April-May and an autumnal pulse of less regular occurrence, location, and size. The optimum temperature for their appear- ance in the plankton lies below 70°, and in winter months they occur but rarely. DISCUSSION OF SPECIES OF CONJUGATE. Closterivim acerosmn Ehrbg.—Average number, 348. More than three times as abundant in the previous year. This desmid is perennial in the plankton, having been found in every month of the year, but at irregular intervals, and never in large numbers. Its distribution is such as to suggest that it is at the most only semi- limnetic in habit. The numbers are too small to follow closely the seasonal distribution. There are pulses on May 3 (3,200), Septem- ber 6 (2,400), and November 1 (2,500) in 1898; and in 1897 a pulse on June 28 (2,000) and one on September 21 (24,000). In previ- ous years vernal pulses in April and occasional autumnal pulses are to be noted. In so far as the optimum temperature is indicated, it 61 seems not to lie near either e>ctreme, and above rather than below the average for the year. Closterium gracile Breb.*—Average number, 49,616 (silk, 305). rhis species was found in small numbers from March to December, ^nd shows pulses on May 17 (1,600) and September 27 (6,400) at temperatures of 64° and 73°. The tenuity of the form of the frustule of this species suggests a limnetic habit. Closterium htnula Ehrbg.—Average number, 556. This also is a perennial species, and is somewhat more abundant and constant than C. acerosum. It likewise has a vernal pulse, which in 1895 appears on April 29 (2,915) ; in 1896, on May 1 (5,364) ; in 1897, on May 25 (3,200); and in 1898, on May 24 (6,000). In both this species and C. acerosum there are slight indications of recurrent minor pulses which are often coincident in the two species. Nine buch movements appear in 1898. The autumnal pulses are less regular in their appearance and size than the vernal, and appear from September to November. The optimum temperatures seem to He between 45° and 70°. This species is only semi-limnetic, and never attains the fluctuations which characterize most limnetic Drganisms. Doubtless other so-called species of Closterium have been included among the variable organisms referred here to C. lumda and C. acerosum. Cosmarmm constrictmn Delp. was found occasionally from March to September, and is probably adventitious. Cosmocladium saxonicum De By.—A single isolated pulse of this minute limnetic desmid appeared in the filter collections of Septem- ber, 1897. It was first noted on August 31 and disappeared after September 29, and was never found at other times in the plankton. The pulse culminated September 9 at 13,500,000*. Gonatozygon hrehissonii De By.—The filaments of this desmid ^ere noted in the plankton only in March, 1899, attaining a maxi- mum of 136,800 on the 14th. Staiirastrum gracile Ralfs.-—Average number, 31. About two hundred times as abundant in the plankton of 1897. It occurs from March to January. No vernal pulse was detected, but an autumnal Dne of 14,000 appears September 29. It appears in much larger numbers in the filter-paper collections, and is probably a limnetic planktont in our waters. 62 Undetermined species of Penium, Arthrodesmiis, and Docidium ha\'e been found in the plankton but always singly. They are doubtless adventitious. PHANEROGAMIA. The Lemnacecu are represented in our waters by several species of Lemna, by Spirodela, and by two species of Wolffia—hrasiliensis and Columbiana. The first two genera are predominantly floating surface-plants, while the last occurs at all levels, is taken with the plankton, and has been treated in our measurements and enu- merations as a limnetic organism. Wolffia hrasiliensis Weddell.—Average number, 2 ; in 1897, 13. It appears irregularly in river planktons from the last of March till January, and is somewhat more abundant in late summer and autumn. The seining operations of fishermen in the river and tributary backwaters have much to do with its appearance in the plankton of the river. Wolffia Columbiana Karsten.—Average number, 7; in 1897, 41. With the preceding species. Neither of these species are sufficiently abundant greatly to affect the ecology or quantity of the plankton of the river, though they are of more importance in the backwaters. Owing to their size and duration they compete with the smaller organisms of the phytoplankton, but do not serve as food for any of the zooplankton. pjRp t[o;zjo;a!. Average number, 111,731,000. The number of species exceeds 147 ( + 38), distributed as follows: Mastigophora, 60(+10); Rhizop- oda, 31 ( + 28) ; Hcliozoa, 5 ; Sporozoa (3) ; Ciliata, 45 ; and Suctoria, 5,—the numbers in parentheses indicating the additional forms whose specific rank was not recognized in the enumerations . The Protozoa occur in great numbers (Table I.) in every collection of the year. Owing to the fact that the totals are a conglomerate of two methods of collecting, of a large number of species of many di- vergent seasonal tendencies, and of both euhmnetic and adventitious forms, their seasonal fluctuations have no particular significance which is not better treated either in connection with the subdivisions of the class or with the individual species. In the totals, traces appear of 63 the vernal pulse, of the midsummer maximum of the chlorophyll- bearing Mastigophora, and of the autumnal-winter wave of Ciliata. The Protozoa, through the Mastigophora, share with the algas the synthetic function in the elaboration of food from inorganic or partially disorganized organic contents of the water. They utilize decaying organic matter as food, and are thus primary links in the cycle of food relations. Some of them feed upon bacteria, upon algae, or even upon other animals, and thus become secondary or tertiary links in the chain. MASTIGOPHORA. (Plates I. and II.) Average number, including, without duplication, both silk and filter-paper collections, 95,856,449. In the collections of 1897 they were five times as abundant as a result, in part at least, of the extended low-water period, sewage contamination, and extension of high temperatures during the late autumn of that year (Pt. I., PI. XL). The Mastigophora abound in every collection and occur at all seasons of the year. Four fifths of them occur, how^ever, between the first of April and the last of September. They are predominant- ly chlorophyll-bearing organisms, and have their greatest numbers during the same season in which the land flora attains its growth. They spring into abundance with the opening buds of April, and van- ish from the plankton when frost cuts off the foliage in autumn. There are, it is true, some species, such as Syniira, which grow luxuriantly at winter temperatures, but these are generally of the chrysomonad type, with yellowish or brownish chromoplasts. The bright green chlorophyll-bearing flagellates are in the main summer planktonts. Since \vater temperatures do not fall below 32°, the phytoplankton is exempt from this risk of destruction against which the land flora must provide. We find, accordingly, that the most of the Mastigophora are wont to occur in diminished numbers and irregu- larly in the plankton throughout the winter. This appears in the records of the more common species, and fuller examination would- doubtless greatly increase the number which thus winter over in reduced numbers. I have already called attention to the fact that there are in 1898-99 recurrent pulses in the ChlorophycecE and BacillariacecB at w > i5 o g J I— I w ^ 66 intervals of several weeks, and that such pulses can also be traced back into 1897 as far as the collections were made at weekly inter- vals—that is to the^early part of July. A similar periodicity on the part of the Mastigophora—the greater part of which are also chlorophyll-bearing—is even more evident. Not only is this periodicity present in this group, but it coincides approximately in the location of its maxima and in their relative development with that found in the ChlorophycecE and BacillariacecB. The following table, which gives the dates of culmination of the pulses of these three groups from July 1, 1897, to April 1, 1899, will serve to demonstrate this point more clearly, and a graphic presentation of the data will be found in Plates I. and II. There are twenty-two of these recurrent pulses in the period from July, 1897, to March, 1899. Of the sixty-six possible maxima only five are missing, or at least not apparent in our data, and but ten culminate on other dates than the one (of collection) most to be expected. These ten in every case culminate either a week prior or subsequent to that in which the other two groups reach their max- ima. These divergences may be due to the error incident to the interval of collection, and their approximation in time is still cor- roborative of the tendency towards recurrent periods of growth. These exceptions are no greater than might be expected to occur in the unstable fluviatile environment and within the large margin of error of the plankton method. There are twenty-one intervals betw^een July 14, 1897, and March 14, 1899, with a range in length of 20 to 42 days and an aver- age of 28.95. The intervals in days with the numbers of instances of each are as follows: 20 (1), 21 (3), 22 (1), 23 (1), 26 (1), 27 (1), 28 (7), 35 (3), and 42 (3), days. The effect of the weekly interval of collection is seen in the preponderances at 21, 28, 35, and perhaps at 42, days. There is evidently a tendency towards the interval of 28 days. Nine of the 21 pulses are grouped about this interval ; 6, about that of 2 1 ; while 3 are at 35 and 3 at 42. If there be such a tendency it is but natural that with a weekly interval of collection there should also appear minor preponderances at 2 1 and 35 days. Traces of a similar rh3^thm may be found in the period of weekly collections in 1896 (Pt. I., Table III.). In some instances the environmental conditions at these times of departure are such as to suggest that they may have produced the 67 shifting in the position of the maxima. Thus the pulse of January 25. 1898, appears after a 3 5 -day interval, but in the midst of the rising winter flood, to whose effect the delay may be attributed. In both 1896 and 1898 the 28-day rhythm is interrupted at the time of the vernal pulse in April-May. It appears as though these re- current pulses—if such exist—Vv'ere submerged in the greater ver- nal increase. The double summit of the vernal pulse in the curve of the BacillariacecE and Mastigophora (PI. II.) for 1898 suggests the compound character of this pulse in the case of these groups of organisms at least. The time interval in the case of the vernal in- terruption is also significant. In 1898 there are two pulses between March 22 and July 19, at intervals of 42 days—a total of 84 days, which is the equivalent in duration of three 28-day intervals. The total number of species of Mastigophora recorded by me from the plankton of the Illinois River is over sixty. This number will be increased to more than seventy if forms not separated in our enumerations be distinguished as separate species. The Protomastigina (including the BicoscecidcB and the Ci'as- pedomonadidce) are well represented in the plankton by passive limnetic species which are principally sessile on other planktonts. These are Bicosoeca lacustrts, SalpingoBca hrunnea, S. minuta, and Diplosiga frequentissinia. Asterosiga radiata is a eulimnetic repre- sentative and Anthophysa vegetans an adventitious one. As a group they are more abundant during the warmer part of the year. The ChrysomonadidcE are also well represented, and include the most abundant flagellates of the plankton of the colder months. Symira iivella is quantitatively the largest factor furnished by this group. It is supplemented by Syncrypta volvox, and the various forms of Dinobryon, Uroglena, and Mallomonas. The last two genera have more of a summer range of occurrence, but are not of quantitative importance in the waters of the Illinois. The Cryptomonadidce are represented only by Chilomonas and Cryptomonas, and are of somewhat constant, though of minor, importance quantitatively. The Euglenidce, on the other hand, are, in our waters at least, second to no coordinate group in their quantitative importance. They are individually of relatively large size, and they occur in great numbers throughout the summer months, replacing the ChrysomonadidcE of the colder seasons of the year. Euglena (6) 68 viridis is the most abundant, and it is associated with other species of the genus, with species of Amhlyophis, Phacus, Lepocindis, Chlo- ropeltis, Colacium, and Trachelomonas, especially the latter. The PeridimidcB are quantitatively of considerable importance in the plankton of our Great Lakes (Kofoid, '95), but in the Illinois River they are of little significance, at least the larger forms such as Ceratium. Smaller species such as Peridinium tabulatum and Glenodinium cinctum are more abundant. As a group they do not show any marked seasonal preferences. The VolvocidcB, on the other hand, are of more than the usual consequence in the plankton of the Illinois. The group is repre- sented by the envious Chloraster gyrans, by the sporadic and meteor- ic Carteria midtifilis, and by the colonial genera Eudorina, Pando- rina, Pleodorina, Platydorina, and Volvox. As a group they are almost exclusively summer planktonts. The Mastigophora as a whole are, next to the BacillariacecB, the most abundant of the synthetic organisms of the plankton. Their quantitative importance has not hitherto been sufficiently demon- strated in the plankton of fresh water, owing it may be to their escape through the silk net in the ordinary methods of collection. It seems quite probable also that they may be present in our warm and fertile waters in much greater abundance than they are in the colder and clearer waters of most lakes. This is especially true of the EuglenidcE and VolvocidcB, perhaps less so of the Chrysomonadidas and PeridiniidcB. I DISCUSSION OF SPECIES OF MASTIGOPHORA. Amhlyophis viridis Ehrbg.*—Average number, 63,014 in 1897. It occurred throughout the summer in 1897, from May to October, with a maximum of 1,440,000 on August 31. Apparently a sum- mer planktont but never very abundant. Anthophysa vegetans (O. F. Miill.) Blitschli.—This was identi- fied in the plankton of June, 1898. It is very abundant at times on various substrata in stagnating water, and from such places becomes adventitious in detached fragments of colonies in the plankton. Asterosiga radiata Zach.—This interesting colonial and limnetic choanoflagellate, described originally from the plankton of German lakes, has been found but a single time in our plankton—in the latter 69 part of August, 1896. It is one of many illustrations of the cosmo- politan distribution of plankton organisms. Bicososca lacustris J. Clark*.—Average number, 1 1 2,896. Onlyone third as abundant in 1896, and four times as many in 1897. This minute flagellate is found in our waters sessile upon the filaments of Melosira, principally M. granulata var. spinosa. It occurs more frequently upon the dead frustules than upon live ones, and upon those of the shorter form than upon the longer. It has appeared also upon Dinobryon sertitlaria, Pediastriim pertusum, and Richteri- ella hotryoides. It exhibits a considerable range of variation in proportions, in the amount of lateral compression, and in the length of the pedicels. These variable forms are, however, connected with the type as described by Clark, and are not, in my opinion, to be designated as distinct species. Zacharias ('94) has described one of these variants as B. oculata. I regard it as a growth condi- tion of B. lacustris, and not as specifically distinct from it. Its seasonal distribution in 1898 is somewhat peculiar. It appears as two quite symmetrical pulses, the first extending from early in June till the middle of July, and culminating on June 14 at 3.801,600. The approach of this pulse is abrupt and its decline somewhat gradual. The species does not reappear until September 13. The autumnal pulse culminates October 11 at 486,000, then gradually declines, and disappears November 1 . There is no record of its occurrence in 1898 outside of these two pulses. In 1897 it is found irregularly from May to August, and in 1896 in February and from May to December, with pulses in May, June, July (2), August, and October. In 1898 its optimum temperatures appear at 82° and 65°, and its pulses in other years do not occur below 57°. It thus belongs to the plankton of the warmer months. Its seasonal distribution falls within that of the limits of its host Melosira, and in 1896 and 1898 their vernal pulses coincide, and the same correlation appears in all but one of the pulses of 1896. Not all Melosira pulses, however, are attended by an increase in Bico- scBca. Thus in the late summer and fall of 1897 Melosira fluctuated without any appearance of Bicosoeca. In the autumn of 1898 the pulse of Bicosaeca on October 1 1 appears on the decline of the Sep- tember pulse of Melosira, in which the host made no corresponding increase. Melosira is thus apparently essential for any marked in- 70 crease of Bicosceca in the plankton, but is not in itself the primary- cause for its appearance in the plankton. Carteria miiltifilis (Fres.) Dill.*—Average number, 2,365,384. In 1897 more than one hundred-fold as abundant. This species was recognized only in the autumnal and hiemal planktons, from August till January in 1897-98 and from October to February in 1898-99. It is not easily and with certainty identified by the usual methods of plankton counting, and probably other species of similar habitus may have been included to some extent ; and, on the other hand, many Carteria may have been thrown with the "un- identified" flagellates, especially in earlier years. This species occurs throughout the whole range of temperatures, and its maxi- mum development (6,476,400,000) w^as attained October 5, 1897, at 70°. A pulse prior to this appeared September 7, at 2,846,250,000. From the major pulse in October there is a gradual decline as the minimum temperatures are reached. The remarkable outbreak of Carteria in the autumn of 1897 was associated with unusually low water (Pt. I., PI. XL) and concen- tration of sewage and decrease in current. The water of the stream was of a livid greenish-yellow tinge, due principally to great numbers of Carteria, which developed to the exclusion or diminution of other chlorophyll-bearing flagellates such as Euglena, and of diatoms such as Melosira. This unusual development seems to have been a dis- turbing factor in the usual seasonal routine of the autumnal plank- ton of that year. The distribution of Carteria in the river was remarkable. It formed great bands or streaks visible near the surface, or masses which in form simulated cloud effects. The distribution was plainly uneven, giving a banded or mottled appearance to the stream. The bands, 10 to 50 meters in width, ran with the channel or current, and their position and form were plainly influenced by these factors. No cause was apparent for the mottled regions. This phenomenon stands in somewhat sharp contrast to the distri- bution of the usual water-bloom upon the river, which is generally composed largely of Euglena. This presents a much more uniform distribution, and unlike the Carteria is plainly visible only when it is accumulated as a superficial scum or film. Carteria was present in such quantity that its distribution was evident at lower levels so far as the turbidity would permit it to be seen. It afforded a 71 striking instance of mfirked inequalities in distribution within small areas, of at least one plankton organism. Carteria showed great variation in the amount of chlorophyll present. Some individuals were practically colorless. It seems very probable that in the presence of great abundance of partially decayed organic matter such as occurs in a sew^age-laden stream, Carteria may become largely holozoic in its nutrition, as Zumstein ('99) has shown to be the case with Euglena. The literature of fresh-water plankton contains no record of a similar preponderance of Carteria in other localities, though its occurrence has been occa- sionally noted in the plankton. The chemical conditions under which this great pulse of Carteria appeared in the autumn of 1897 can be foliow^ed in Part I., Plate XLIV and Table X. The high chlorine and the great increase in free ammonia and nitrites indicate the decay of sewage; the high nitrates and albuminoid ammonia show that there was no lack of some at least of the important sources of foo.d- The two principal pulses appear September 7 (2,846,250,000) and October 5 (6,476,- 400,000), with a minimum of 680,400,000, on September 21, sep- arating them. Both of these pulses are attended by sharp declines in nitrates and nitrites and free ammonia, and very slight decreases in organic nitrogen and albuminoid ammonia. Either the first three substances named or those matters which supply them by their decay, are thus noticeably utilized at the times of these pulses. The relation of the Carteria to the volumetric pulses is (Pt. I., PI. XI.) not a constant one. The Carteria pulse of September 7 lies in a slight depression betw^een two maxima of the volumetric curve, and a w^eek prior to the autumnal culmination on September 14 at 19.8 cm.^ per m.^ It thus appears during the growth period of this volumetric maximum. The second and larger pulse of Carteria, on October 5, coincides with the second volumetric maxi- mum, and in fact fluctuates throughout w4th it. Though Carteria constitutes but a small part of the actual catch of the silk net, owing to leakage through the silk, it is apparently an important factor in the food cycle which builds up such maxima. Ceratium brevicorne Hempel.—This species appeared in small numbers in isolated instances from April through October. It varies towards C. hirundinella, but the small numbers in which it has occurred have not as yet afforded sufficient ground for regard- 72- ing it as a variety of that species. It occurs most frequently in August and September, and is apparently a warm-water planktont. Ceratium cornutum Ehrbg. was found but once—in June, 1896. Ceratium hirundinella O. F. Miill. was not noted in our plank- ton in 1898, but in 1896 was found from June to October, with a pulse of 19,200 on June 6. It was recorded only at temperatures above 57°, and is apparently a warm-water planktont. It has but an insignificant part in the potamoplankton of the Illinois River and its backwaters, though quite abundant in the summer plank- ton of Lake Michigan (Kofoid, '95). It seems not to have survived the transit through the sewage-laden waters of Chicago River or to thrive in the conditions prevailing in the Illinois River, though common generally in fresh-water plankton of the temperate zone. Chilomonas paramcecimn Ehrbg.*—Average number, 555,000. This flagellate, which is frequently abundant in aquaria or stag- nant water, appears also in the plankton of the Illinois River. There is in 1898 a vernal pulse, culminating at 10,800,000 on April 26, and there are scattered records from October to February. Chloraster gyrans Ehrbg.—This rare and unique flagellate was found in but two collections—in July and August, 1898—and only in small numbers. Chloropeltis monilata Stokes.*—Average number, 362,941 in 1897. This is a summer planktont, appearing at irregular inter- vals from the last of May until the middle of September. It was not found in 1898. A maximum of 10,800,000 appears oh August 31. Colacium calvum Stein.—The attached stage only of this flagel- latewas observed , and was recorded only in 1 896 and 1897. It appears from the middle of April to the first of October, and is usually found upon Polyarthra platyptera. It has occurred occasionally upon several species of Brachionus and upon Chydorus sphcuricus. The largest number recorded (162,792) appeared on April 17, 1896, upon Polyarthra, usualty upon the body and more rarely upon the oar-like appendages. It is often exceedingly abundant upon the planktonts of backwater ponds. Colacium vesiculosum Ehrbg.—This species is much less abun- dant than the preceding species in our waters, and was found only in June and September, upon Cyclops alhidus and Polyarthra. 73 Cryptomonas ovata Ehrbg.*—Average number, 121,154. This species has been recorded principally in the autumnal or hiemal plankton. It escapes through the silk net readily, and was rarely found in collections of earlier years. In 1895 it occurred from July till the last of October, and in 1898 was common in the December plankton. Dinohryon sertularia Ehrbg.—Like most typical pkinktonts, Dinohryon is an exceedingly variable organism, and the varia- tion finds its expression in the form and proportions of the loricas and in their arrangement and continuity in colonies. Divergences from described and figured species are thus at once apparent, and they have been utilized by systematists, notably by Lemmermann ('00) and by Brunnthaler ('01) as the basis for the establishment of a large number of new species. The validity of these species, in my opinion, must rest ultimately upon careful experimental evidence of their present mutual genetic independence under normal conditions of growth. From my own observations upon large numbers of colonies and individuals distributed through- out the range of their seasonal recurrence in six years in our waters, I am inclined to regard all as belonging to a single species, and the different types as mere growth varieties. The rapidity of growth and the age of the individual or of the colony are, I believe, impor- tant factors in the determination of the form of the lorica, and its various forms are therefore not of specific value, but rather of physiological significance. It is a simple matter to find individuals, or even colonies, conforming to the descriptions of the several species, but it is not so easy to refer all individuals and all colonies to the described types. They intergrade—nay, more, two, or even more, "species" are not infrequently combined in the same colony. I have never found all the forms in a single colony, but such com- binations as angulatum-divergens, divergens-angulatum-sttpitatum, sertularia-angulatum, and sertularia-undulatum have been observed by me. These combinations are most frequent in large colonies, and, indeed, the number of "species" in a colony is apparently a function of its size. The slender growing tips are wont to assume the stipitatum type of lorica and colony, and the older loricas at the base to conform to that of sertularia, divergens, or angidatum. Small colonies as a rule belong to a single "species." These com- binations are generally most evident during the maximum period of growth ; that is, when Dinohryon is multiplying rapidly, though they may appear at any season of its occurrence. In the enumeration of Dinohryon five types were recognized, and the individuals were assorted to these "species," viz. : D. sertularia, stipitatum, divergens, angtilatum, and undulatum. Some corrobora- tion of the view that we are dealing with a single variable organism and not with five distinct species may be seen in the coincidence of the seasonal distribution, and of the rise, culmination, and decline of the pulses of the five different forms. Since these varieties have such a similar seasonal distribution I shall treat them as a w^hole, discussing subsequently any individual peculiarities which are noteworthy. The average number of indi\'iduals of Diyiohryon sertularia, including all its varieties, in 1898 was 1,979,785. In 1897 the average was much smaller (79,352) owing to the few collections in the winter, when it is most abundant, and to its suppression in the prolonged low water of the autumn of that year. The relative frequency of these different varieties—for I shall treat them as such—is shown by the average per cubic meter for the year in 1898, viz.: D. sertularia, 407,602; D. sertiUaria var. stipitatwm, 603,911; D. sertularia var. divergens, 866,083; D. sertularia var. angulatum, 101,358; D. sertularia var. undulatum, 831. These figures are only approximate, since colonies containing more than one variety have all been included with the predominant variety in the colony, which is usually sertidaria or divergens, consequently angulatum and undulatum are more numerous than indicated by these figures. The seasonal distribution of Dinohryon in our waters is well defined, and is sharply limited to the period from November to June. Its earliest recorded appearance was November 8 in 1898, while in 1896 and 1897 it w^as not found until in December. It lingers well into June in 1896 and 1898—the two years in which the spring collections were of sufficient frequency to trace its decline. In 1898 the latest record was on June 28. Most of the records after May are irregular and sporadic. It is thus absent from the plankton of the Illinois River from the last of June till November or December. In 1895-1896 there was also a winter interval in which no Dino- hryon was recorded during the December-January flood (Pt. I., PI. IX. and X.). In 1897-1898 a similar interval appears, and con- tinues almost to the end of the slow^ rise of the flood which culmi- 75 natecl in March. Rising floods thus do not favor the development of Dinobryon in channel waters of the Illinois. The interval of collection in 1894-95 is too great to trace the seasonal fluctuations of Dinobryon, though there are indications of a maximum pulse on April 29. In 1895-96 there is a slight de- velopment in November prior to the rise of December, in which Dinobryon again disappears. A slight pulse of 3,192 appears on the declining flood (Pt. I., PI. X.) on January 25, and declines again with the rise in February to reappear on February 20 at 42,588. Another decline in Dinobryon attends the rise in river levels in February-March, and after a fortnight of falling levels a third pulse of 2,531,280 is seen on March 17. Two other pulses attend the decline of this flood, one upon April 29 (800,064) and the other on May 18 (339,624). On the decline of the June rise of this year a late and unusually large pulse for the season appears (June 11) at 2,438,400. An examination of the hydrograph will indicate that almost without exception these pulses attend the run-off of im- pounded backw^aters after recent invasion, or, as on April 29 and May 18, after a temporary check in the run-off. During those times when the channel contributes to the backwaters, that is, dur- ing rising floods, Dinobryon declines in numbers; and, on the other hand, it reaches its greatest development in channel waters during the run-off of the flood. In 1896-L897 the interval of collection (Pt. I., Table III.) is again too great to trace satisfactorily the fluctuations of Dinobryon. There is a pulse on December 3 of 157,609 and on April 27 of 172,800. In {897-98 Dinobryon appears first on December 7, with a pulse of 1,807,200, during a period of low water and ice blockade with no backwater contributions. It declines, and after December 21 does not again return until March 22, when an isolated record appears. The vernal pulse begins April 19 and culminates May 10 at 84,- 841,600 on the declining spring flood (Pt. I., PI. XII.). Dinobryon declines at once during a fortnight of rising water, and t\vo minor pulses on the decline of the flood—one on June 7 of 70,400 and one on June 28 of 219,840—complete its vernal cycle. The hydrographic conditions in 1898-99 were very different from those of the preceding season, and w^e find a marked change in the seasonal occurrence of Dinobryon. From November to March 76 there are three rises to overflow stages (Pt. I., PI. XII. and XIII.) with intervening decHnes of a month's duration. There is a pulse of Dinohryon in each of these periods of dechning flood. The pulse of 275,200 on December 20 follows the November flood, and it is foUow^ed by a minimum of 1,500 on the rising flood of January 10. The numbers slowly increase until a meteoric rise on February 7 to 6,486,700 and on February 14 to 22,621,440 is followed again by another decline, to 25,920 on February 28, with the sudden flood of that week. During the maximum flood stage in March (Pt. I., PL XIII.) no Dinohryon was recorded, but it reappeared again on March 21. The suspension of our plankton operations interrupted the further tracing of the fluctuations. From the facts above detailed it is very evident that the pulses of Dinohryon occur in channel waters at times when the run-off of impounded backwaters is making its greatest contribution to the river plankton. These are times of greatest stability of the en- vironment in all respects save river level and its sequences. The impounded waters have come from regions of slight current and decaying vegetation, and there has been time in those localities for the decay of sew^age and debris, and for the growth of planktonts such as Dinohryon. These conditions of the environment are therefore favorable for the growth pulses of Dinohryon. The phenomenon of pulses of growth is not, however, to be considered as merely the result of declining floods. These afford a favorable environment and doubtless determine within certain limits the time and the extent of the pulse. The phenomenon is one common to most plankton organisms, and occurs in Dinohryon of lakes where floods are of little significance. Any evidence of recurrent minor pulses in Dinohryon at brief intervals is lacking. Dinohryon has been found in our plankton through practically the whole range of temperatures, but it disappears when maximum summer heat is reached and does not return until the water cools to 45° or lower. Large pulses, such as that of February 21, 1899 (22,621,440), have developed at temperatures approximating 32°, and largely under the ice. The vernal pulse of April-May has been recorded at temperatures ranging from 60° to 79°, but generally nearer the former. No well-defined optimum temperature appears, and the seasonal distribution suggests that the high temperatures 77 of our summer waters are inimical to Dinohryon. That its absence from the plankton at that time is not due merely to low-water con- ditions is shown by the December pulse in 1897, under the most pro- nounced type of such conditions. Dinohryon is a common planktont in the Great Lakes (Kofoid, '95) during the summer months, but surface temperatures here rarely exceed 68°, and are 10° to 20° below those of the Illinois River. In German lakes Apstein ('96) finds the maximum develop- ment of Dinohryon in June and a continuance through the summer in reduced numbers, but temperatures are also 10° to 20° (F.) lower than in our waters. In the case of D. siipitatum there is a second maximum in August. Lauterborn ('93) finds Dinohryon throughout the winter in the plankton of the Rhine, with a maxi- mum in April-May, with diminished numbers during the summer, and a second maximum in September. The filter-paper collections give very much larger numbers, owing partly to the inclusion of small colonies which escape through the meshes of the silk net in the usual method of collec- tion. The numbers are increased at least thirtv-fold if filter col- lections are utilized instead of silk, as above. The size of the colonies in the collections varies greatly, the averages ranging from three to forty-eight cells. The maximum pulse is attended or followed by a considerable decrease in the size of the colony. In the pulse of February 21, 1899, the average num- ber of cells in the colony falls from thirteen to sixteen, during the rise of the pulse, to seven, at its culmination. On the pulse of May 10, 1898, the average is thirteen, and a week later, when the pulse declines from 16,153,600 to 43,200, the average size of the colony drops to three cells. Cysts also are most frequent during and subsequent to maximum development. Dinohryon is some- times covered with large numbers of minute choanoflagellates, probably Salpingceca minuta Kent. Frequently colonies occur in which only the younger cells are alive. Dinohryon is, in the light of its distribution, one of the impor- tant synthetic planktonts of the colder months, and is one of the primal links in the chain of food relations of that season, serving as food for some of the winter Cladocera and Copepoda. The fact that its maxima frequently occur when volumetric minima appear- as, for example, on February 21, 1899—indicates that Dinohryon 78 does not directly contribute much, even at its maximum develop- ment, to the volume of the plankton taken in the silk net. On the other hand, its rapid multiplication, as evidenced by its meteoric pulses, may serve to build up a more permanent and bulkier animal plankton, and thus indirectly, in a cumulative way, it may be of considerable quantitative importance. The inclusion of all the variants of Dinohryon as a single species has been favored by Wesenberg-Lund ('00), who regards D. stipi- tatum as the summer form of D. sertularia. In our plankton, D. stipitatum has occurred sporadically in December and March, but it is most abundant during the vernal pulse in April-May. Its distribution thus in the main supports that author's contention in that it is found during the w^armer portion of the seasonal cycle of Dinohryon in our waters, though not in our summer plankton. It is not desirable in this connection to enter further into a discussion of problems which have been raised by the splitting up of Dinohryon into so large a number of forms. Lemmermann has found seven- teen species and varieties within the limits of the subgenus Eudino- hryon. A discussion of their validity involves not only some per- plexing problems of synonymy, but also an extensive examination of a large amount of material showing seasonal changes, and, above all, a series of experiments which shall demonstrate the limits of variation within a known line of descent and in the sea- sonal range of environmental conditions. It involves, moreover, the fundamental question of the criterion of species. The papers of Lemmermann ('00) and Brunnthaler ('01) have appeared since my work of enumeration was completed. I recognize among the forms which they have sought to establish the following which occur in our plankton: D. sertularia Ehrbg., D. sertularia var. thyrsoideum (Chodat) Lemm., D. sertularia var. alpinum Imhof, D. protuherans Lemm., D. sociale Ehrbg., D. stipitatum Stein, D. stipitatum var. americanum Brunn., D. stipitatum var. havaricum (Imhof) Zach., D. elongatum Imhof, D. elongatum var. undulatum Lemm., D. cylindricum Imhof, D. cylindricum var. palustre Lemm., D. cylindricum var. schauinslandii (Lemm.) Lemm., D. cylindri- cum var. pediforme (Lemm.) Lemm., D. cylindricum var. divergens (Imhof) Lemm., and-D. cylindricum yslv. angulatum (Seligo) Lemm. As a result of my attempts to refer all of the individuals which I have seen in my work of enumeration to species, I am of the opinion 79 that we are dealing in the case of the species of Dinohryon above cited with a single variable organism, whose extremes of variation only have been regarded as separate species. The connecting Hnks are sufficiently abundant still and the union of several types in a single colony is sufficiently frequent to lend some weight to my con- clusions with regard to those forms which have been under my observation. In the interests of utility as well as in the interests of well-grounded taxonomy, it is extremely desirable that the establishment of new species among variable plankton organ- isms should be attempted with extreme caution and only after the fullest study of the range and conditions of variability. The insta- bility of the taxonomic structures which Brunnthaler and Lemmer- mann ha\-e recently raised, is evidenced by the differences in syno- nvmic. varietal, and specific rank given to the variants of Dino- hryon by these two systematists, who have but recently mono- graphed the group, largely if not wholly from the systematic point of view. The changing estimate of validity w^hich Lemmermann himself has put upon his own species or varieties—for example, schauinslandii, pcdiforme, and curvatum—gives further evidence that the basis upon which they rest is at the best biit slight. It is my firm conviction that the establishment of new species among the organisms of the plankton of fresh water can be satisfactorily accomplished only after careful analysis of the limits of variation within the range of environmental conditions. Stand- ards less comprehensive than this can yield results of but temporary or local value and can lead to but little permanent advance in science, and they bring only perplexit)" and chaos where order should reign. Diplosiga jrequentissima Zach.*—Average number, 1,736,538. This minute flagellate is found upon the rays of the colonial diatom Asterionella, often in great numbers and so thickly set as to leave little unoccupied space. It was found in each year at the time of the vernal pulse of Asterionella in April-May, and was as a rule most abundant immediately after the maximum growth of Asterionella had been attained. Beyond an isolated occurrence in January it was not recorded at other times than during the months of April and May. Eudorina elegans Ehrbg.—Average number, 14,362. About twice as abundant in 1897. The distribution of this species is 80 somewhat erratic. It has occurred in every month from February through October, but in smaller numbers and sporadically in the colder months. In 1898 its seasonal curve is of characteristic form. It makes its appearance March 1 5 , and is continuously'present until the end of September. There is a vernal maximum April 26 of 240,000, but no corresponding autumnal one. In 1898 there are indications of recurrent pulses at brief intervals which coincide in location immediately or approximately with similar ones of Gonium and Pandorina. These pulses occur March 15 (3,600), April 5 (2,800), April 26 (240,000), June 14 (60,000), August 2 (8,000), August 23 (3,200), and September 20 (2,000). The minima between these pulses in all cases but one fall below 1,000. In 1897 a vernal pulse was not detected, a maximum of 496,000 occurring August 3 1 , and but three minor pulses appearing. In 1896 this species appeared in the plankton on February 20, and remained until the end of August with a month's interrup- tion in May-June. There were no marked pulses, exceeding 15,000, in that year. The absence of the spring flood (Pt. I., PI. X.) and the disturbed hydrograph of the summer may account for this suppression of development in Eiidorina. The distribu- tion in preceding years is also irregular. Eudorinahegms, its seasonal development at temperatures but slightly above 32°, but any considerable growth is not attained until at least 45° has been reached, and the largest pulses on record have been at the close of the period of maximum summer heat at a temperature of 80°, and the vernal pulses have been at 60° or above. The disappearance of Eudorina from the plankton in the early fall, about the time that foliage is killed by autumnal frosts, has been constant in the different years. Eudorina is not sufficiently abundant to be of any considerable importance in determining directly the volume of the plankton. It serves as food for many of the rotifers, and is itself frequently parasitized by Dangeardia mammillata Schroder, which destroys the cells but leaves the matrix intact. There are times when it is hardly possible to find perfect colonies, and when it is not unusual to see colonies swimming about propelled by one or two surviving cells. Euglena acus Ehrbg.*—Average number, 214,807. Found from the middle of March till the first of November, and most abundantly 81 in late summer and early autumn. It escapes through the silk net readily, and no marked pulses in occurrence appear in the erratic data of the filter-paper collections. It is found in the water-bloom, and is predominantly a warm-water planktont. Englena deses Ehrbg.—Occurs occasionally in the plankton and water-bloom during summer months. Englena elongata Schew.*—Average number in 1897, 278,970. It is found irregularh^ in our plankton and water-bloom from July to October. Originally described from New Zealand. Englena oxynris Schmiarda.*—Average number, 960,769. Next to E. viridis this is the most abundant member of the genus in our plankton. It is abundant during the summer, especially towards its close during low-water conditions, when the water-bloom, in whose formation it shares, is best developed. There is no vernal development, and the fluctuations are but slight in com- parison with those of most organisms of the plankton. There is a slight indication of recurrent pulses at intervals of a few weeks. Its optimum temperature lies near that of maximum summer heat, that is, about 80°, though some tendency to run over into autumn months is manifest. Englena sanguinea Ehrbg.—There are only sporadic occurrences of this species in the plankton. It is found along with E. viridis among matted growths of LemnacecE, and on exposed and reeking mud flats, where it forms patches of bright red color often of large extent. It may be only a physiological condition of E. viridis, with which it is always found. It has appeared in the plankton most frequently in September, though found elsewhere throughout the summer. Euglena spirogyra Ehrbg.—Found but once—in October, in the river plankton. Euglena viridis Ehrbg.*—Average number, 1,571,731 ; from silk collections only 8,653. This is the most abundant of the larger green flagellates in our plankton, and constitutes the greater part of the water-bloom of summer months, when it forms towards four p. m. a livid green scum on the immediate surface of the water. Collections of the silk net give no clue to its abundance and shed no light on its seasonal distribution. The filter-paper collections indi- cate its presence from March to December, but in numbers only during the warmer period, from May to October. There is no ver- 82 nal pulse though there are shght traces of minor irregularities, and on September 7, 1897, a single unusual deyelopment of 58,000,000. Its optimum temperatures lie close to the maximum heat of summer months. It is found not only in water-bloom and plankton, but also along shores, on mud banks, and in sequestered pools and bays where temperatures reach 90° and over. Lightly colored and semi- transparent individuals of this and other species of the genus are found frequently in the plankton, suggesting an approach to holo- zoic nutrition in nature, such as Zumistein ('99) has demonstrated experimentally in E. gracilis. Euglena is quantitatively one of the most important links in the chain of food relations of the summer plankton, converting nutrient matters in the water, both organic and inorganic, into food for the Rotifera and Entomostraca of that season of the year. It in a measure replaces the diatoms, some of which decrease in number or disappear during the warmer months. Glenodinium cinctum Ehrbg.*—Average number, 1,360,192. This species is generally present from the middle of March till the end of September, though sporadic occurrences are found in winter months. There is a pulse on March 29 of 4,260,000 at a temperature of 49°, and another August 9 of 25,200,000 at 83°. This small planktont usually escapes through the silk net. It may be that several species have been included, as the conditions of plankton enumeration do not permit close scrutiny of such small organisms, lacking prominent structural characteristics. It seems to be a perennial planktont with a wide range of temperature adaptation, and with a growing period approximating that of the land flora of our latitude. Gonium pectorale O. F. Miill.—This colonial flagellate has been found in the water-bloom in large numbers, especially in the back- waters. It was taken in the river plankton in 1897 and 1898 in May and again in August and September. These pulses coincide in location with those of Pandorina and Eiidortna. Lepocinclis ovum Ehrbg.*—Average number, 401,538; silk 3,719. This species appears in the plankton in April and continues until the end of October, with sporadic appearances in winter months. There is no vernal pulse, and in both 1897 and 1898 maxi- mum numbers, 43,200 and 50,400, occur at the height of midsummer heat in August. In both years there are well-defined recurrent pulses at intervals of three to six weeks to be traced in the silk 83 collections. The optimum temperatures plainly lie near the maxi- mum, that is, about 80°, and the season of growth approximates that of the land flora, being limited to the months of April-Septem- ber. This is a variable organism, and a number of species have been described in the genus in recent years. Many of these occur in our waters, but no attempt has been made to separate them, since they are based on minute characters. Mallomonas plosslii Perty. and M. prodncta Zach.—These two forms ^^'ill be treated together, as in my opinion they are merely divergent variants—perhaps seasonal—of a single species. In 1898 M. plosslii was found but three times—in June and July—and M. prodncta eight times—from ]\Iay through September. In 1897 the latter only was recorded, and in September and October. In 1896 M. plosslii appeared in July and M. produda in April and August. In 1895 M. prodncta alone was recorded, and that in November. The data are hardly sufficient for generalization, but so far as they go they indicate that prodncta is more prevalent in late summer and autumn and pldsslii in early summer, the more attenuate form {prodncta) in the warmer season. Butschli ('80-'89) has intimated that there may be some genetic connection between Mallomonas and Synura uvclla. Certain features of its occurrence in our plankton lend their support to this view. Synura in our waters is a \^'inter planktont, with December and February or March pulses. Mallomonas is a summer planktont, making its first appearance during the time of the decline of Synnra, and when many of the colonies of the latter are breaking up into their individual zooids. Again, the differences in structure and size between the t\\'o genera are quite superficial, and inight result from the growth attending the free life of a Synnra zooid and its prepara- tion for sporulation. It is a noticeable phenomenon that the pro- portion of sporulating individuals of Mallomonas in the plankton is exceptionally large among all plankton organisms. "Free cells" of Synnra are plainly referable to that genus by their resemblance, and by the fact that they are often united in clusters of several indi- viduals forming fragments of disintegrating colonies. It may be that some reproductive phase, as conjugation, intervenes between the free-cell condition of Synura and the Mallomonas stage, and that the relati\'ely smaller numbers of the latter are due to the in- frequency of this process. While the features of seasonal distribu- (7) 84 tion, structure, and sporulation thus suggest the possibiHty that Mallomonas is a free zooid stage leading to sporulation in Synura, they do not demonstrate it, and the genera must stand in statu quo until breeding experiments shall clearly demonstrate the full life- cycle of Synura. Pandorina morum Bory.—Average number, 6,957. In 1898 this organism was about half as abundant as Eudorina, but in 1897 it more than equals it. On account of the small size and the motility of the colonies many of them escape through the silk, so that it is not so adequately represented in silk-net collections as Eudorina. It is probably the most important quantitatively of the Volvocidoz in our plankton. It occurs from April to October, with a few spo- radic appearances in March and up to January. Its greatest growth occurs from May to October. There is no predominant vernal pulse in 1898, but a series of smaller ones culminating May 3 (48,400), June 14 (60,000), July 26 (63,200), and August 30 (3,200),—all upon declining floods (Pt. I., PL XII.) and coincident with pulses of other VolvocidcB—Eudorina and Gonium. In 1897 its seasonal dis- tribution was also similar to that of these genera, exhibiting a max- imum pulse August 31 of 638,000 at 80°. In 1896, a year of inter- rupted hydrograph (Pt. I., PL X.), Panciom^a attained no marked development. Its optimum temperatures lie at and above 60°, and its larger pulses appear during the season of maximum temper- ature, that is, at about 80°. Pandorina does not attain any marked autumnal growth, but declines in September, and as a rule dis- appears in October. The period of its growth thus lies within that of the land flora. As in Eudorina, so also here, parasitism by Dangeardia mammil- lata is of frequent occurrence. Pandorina is an important element in the food of summer rotifers such as Brachionus. Peridinium tabulatum Ehrbg.*—Average number, 3,875,769; silk, 3,711. This is a perennial planktont, having been found in every month of the year. Its principal development is, however, reached during warmer months, from May till September. In 1897 the maximum pulse of 172,800 was on August 10, and in 1898 one of 66,800 fell on July 26, the temperatures being81°and 89° respectively. The only exception to this predominance in warm months is an iso- lated pulse of 2,400 which developed on the declining flood of Febru- ary, 1899 (Pt. I., PL XIII). The absence ofanyautumnal development 85 of this species is noticeable. Its optimum temperatures lie close to the summer maximum (80°), and though perennial, its occurrences at other seasons than late spring and summer are irregular and its numbers few. Its seasonal distribution in German lakes, as re- ported by Apstein ('96), is similar to that in the Illinois River. The PendiniidcB playbut an insignificant part in the plankton of the Illinois River. PJiacus longicaudus Ehrbg.*—Average number, 61,153; silk, 3 ,03 1 . This species in 1898 made its first appearance in the planktonon March 23 and continued till November 15. The species is small enough to escape through the silk net, and the data from such collections do not fully express its seasonal fluctuations. There is no marked \-ernal pulse, and there are traces of but a few small ones during the summer, the largest in 1898 being one of 35,200 on September 27. The distribution in previous years is much the same. A well-sus- tained development throughout the w^armer months—save when rising floods, as that of May, 1898, reduce the numbers—indicates that the optimum temperature for the species approaches the summer maximum (80°). There are almost no occurrences below 45°. This is the most abundant member of the genus in our plank- ton, but it is not quantitatively an important element therein. Phacus pleuronectes Nitzsch.*—Average number, 450,000; silk, 298. It is less abundant (from one fifth to one tenth) than P. longicauda in the catches of the silk net but apparently much more abundant in the filter-paper collections, which may be due in part to its smaller size and greater tendency to escape through the silk in the collections of the net. Its occurrences are even more closely limited to summer months—from June till September. There is no vernal development, and the largest numbers occur during the period of maximum heat. Pulses are but feebly defined. It is also a summer planktont. Phacus pyrum Ehrbg. was found but once—on August 10, 1897. Phacus triqiieter Ehrbg. occurred in small numbers during July and August, 1897. Platydorina caudata Kofoid.—Average number, 17. In 1898 this interesting new genus of the VohocidcB was found in the plank- ton only in the latter part of July. In 1897 it was much more abun- dant (average number, 21,963) and ranged from July 14 to October 86 12. There was a pulse on July 21 of 18,400 and another on vSeptem- ber 7 of 600,000. In previous years the occurrences were scattering, but confined to July, August, and early September. It is evidently a summer planktont, whose optimum temperature lies near the max- imum attained by our waters. No record of occurrence below 60° was made. The smaller and younger colonies escape readily through the silk net. Its pulses in 1897 coincide very closely with those of Gonium, Pandorina, Eudorina, and Pleodorina. Pleodorina californica vShaw.—Average number, 11. In 1897 this species, in common with other members of the family, was much more abundant than in any other year of our work, stable conditions of low water with the accompanying sewage contamination seeming to favor its development. The earliest record for P. californica in the plankton is May 18, 1896, at 71°. This was a year of lower water and higher temperatures than usual in spring months (Pt. I., PI. X.). In other years P. californica did not appear until June or July. It continues into September, the latest record in 1895 being October 2. In 1897 there were pulses on July 21 (5,600) and September 7 (4,000). The occurrences at other seasons are too scattered to trace the seasonal fluctuations, but there is a well-de- fined predominance during the period of maximum heat. This is evidently a summer planktont, whose optimum temperature lies near 80°. Pleodorina illinoisensis Kofoid.—Average number, 6,917 in 1897. This is somewhat more numerous than the preceding species, and its range of occurrences is quite similar. Its maximum pulse in 1897 (180,000) is on August 31, a week earlier than in other members of the family. These pulses of the Volvocidcc occur (Pt. I., PL XLIV.) in a depression of nitrates and just prior to the volumetric pulse of September, 1897. This pulse is doubtless built-up partly at their expense. Their decline in numbers corresponds with its rise. This is also a summer planktont, and was not recorded below 71°. Salpiyigceca brunnea Stokes.*—This species was not recorded in 1898. Average number in 1897, 1,887,356. It occurred on May 2 5 and July 2 1 , dates of culmination of pulses of Melosira granulata var. spinosa. In August-September a pulse occurs, culminating September 7 at 47,250,000—a week after the culmination of a Mel- osira pulse. In 1896 (silk collections only) it was present through- 87 out most of the summer, attending only approximately the sup- pressed and interrupted pulses of Melosira in that year of disturbed hydrograph. It has been recorded from the latter part of April till the middle of September, and, as a rule, above 60°. This beau- tiful little choanoflagellate is sessile upon the filaments of Melosira, principally upon the ^•ariety spinosa, and but rarely upon M. varians or other planktonts such as Pediastrum. It is often associated with Bicosceca lacustris and is usually found upon the sides of the filaments, the bowl of the transparent brownish lorica being closely sessile upon the diatom. In one instance a lorica was found upon the corner at the end of the filament. The lorica had adapted itself to this novel situation by an angular indentation fitted upon the corner of the diatom. Syncrypta volvox Ehrbg.—Average number, 625. This species has a definite and somewhat unusual seasonal distribution. In 1898 it was found from March 1 to April 12, and reappeared Novem- ber 8, attaining a maximum of 13,500 on December 6, and of 43,000 on January 1, declining then to 800 and rising on February 14 to 4,800, and subsequently disappearing in the flood waters of March. It was not recorded in 1897. In 1895 it appeared September 27 and continued for a month, reappearing in February and March, and not occurring after April 10. It has attained its largest develop- ment at minimum temperatures under the ice—43,000 January 3, 1899, at 32.7°. The greater part of its occurrences in 1898-1899 lie very near this temperature, and but three in all the years lie above 50°. It is par excellence a winter planktont, or at least a cold-water one. Its occurrences in 1895-1896 lie near the beginning and the close of the seasonal pulse of Syniira. In 1898-1899 the pulses of Syncrypta coincide in location with or immediately follow those of Synura. The resemblance of Syncrypta to small colonies of Synura is striking, and this fact combined with the relation of their seasonal fluctuations raises the query if Syncrypta may not be an encysting stage of the Synura colony. Its life history should be fully worked out. Synura uvella Ehrbg.—Average number of colonies, 8,463. The seasonal distribution of this chrysomonad flagellate is some- what similar to that of its near relative Syncrypta. It is a perennial, though predominantly cold-water, planktont. It appears 88 in the December plankton of 1894, but was exterminated from the channel plankton taken in the following February by the stagna- tion attending the long-continued ice blockade. It reappears in April, and again disappears promptl)^ but does not return until September 12, and not in numbers until October. There are pulses November 20 (506,800) at 42.8°, and December 30 (362,520) at 36.5°. The December pulse is followed by a decline, with a rise during February to a well-sustained maximum during March, approaching 400,000, and at from 35° to 48°. The decline follows in April, and there are only isolated occurrences in small numbers at irregular intervals during the summer. Continuous occurrence begins again in September, and numbers rise rapidly in October. There is a pulse of 542,699 on December 3 at 32.2°, and another on March 22, 1897, of 159,500 at 43.8°. Symira is very rare indeed in the summer of 1897, and in the prolonged low water, sewage contamiination, and higher temperatures of the unusual autumn of that year it does not reappear continuously until October 26, at 59°, and does not exceed 1,000 until December 7, at 32°. There is a low maximiun of 98,700 on December 14 at 36°, followed by a decline during the rising flood of January-March, 1898. The slight cessations in the flood invasion (Pt. I., PI. XII.) in January and in the second weeks of February and March produce prompt responses in immediate rise in numbers in'Synura. Finally, a low maximum of 320,600 is attained upon the crest of the March flood, on the 29th, at 49°. This is followed by a decline during April and a few scattered appearances during the summer. Synura returns at the end of October and rapidly mounts to a pulse of 1.999,500 on November 29 at 35° with, the first decline of the November overflow (Pt. I., PI. XII.). A second pulse of 2,764,800 on December 20 at 33°, under the ice, gives way to a decline to 51,600 towards the end of January, 1899, during rising water. On February 14 another pulse (348,800) appears at 32.5°, under heavy ice, and declines agaia in the sudden flood of the last days of February, but recovers quickly with a maximum pulse of 898,800 on March 7 at 32.8°. Within a fortnight this falls to the low level of 9,800, but its further history was not followed. From these data it is evident that in our waters at least Symira is limited to the months from October to April, except isolated and irregular occurrences of small numbers during the summer. Its 89 optimum temperatures lie below 50°, and its greatest de\-elopment has taken place in minimum temperatures under the ice. Rising floods and disturbed hydrographic conditions tend to reduce its numbers or to suppress its development, while declining floods initiate increase in numbers and favor the appearance of pulses. A "late" autumn delays the appearance of Synura. Not only are colonies of Synura found in the collections, but at times large numbers of free cells make their appearance. These are released by the breaking up of colonies, and occur in all degrees of isolation. It seems to be a natural phenomenon, and occurs most abundantly with or immediately after the crest of the pulse. Thus the pulse of December 29 (1,999,500 colonies) was attended by 21,600,000* free cells on that date. A week later there were 1,693,500 colonies and 57,600,000 free cells. There are in the rec- ords several instances of meteoric increases of free cells at other times than at those of apparent pulses. It does not seem possible from the data at hand to determine whether this is due to environ- mental influences or to the accidents of collection and subsequent handling. In the discussions of Mallomonas and Syncrypta, sugges- tions have been made that these organisms may be stages in the life cycle of Synura. Synura is the largest and by far the most important S3'nthetic organism of the winter plankton. It shares appreciably in the winter volumetric pulses—as, for example, those of December, 1898 (Pt. I., PI. XII.). Its fluctuations do not seem to produce any marked effect upon the nitrates, possibly because the latter are present in excess of the needs of Synura. In the winter of 1898 nitrates are high, 1.25 parts per million with the pulse of 1,999,500 colonies on November 29, but decline rapidly to .1 on December 13 with a fall of Synura to 78,000. On December 20, Synura rises to 2,764,800, but the ni- trates rise only to .35. It is evident that the nitrates are not the only factor regulating the fluctuations of Synura. ^larsson ('00) reports Synura as abundant in the winter plankton of lakes about Berlin, and Brunnthaler ('00) finds it in the winter plankton of the Danube. There is, however, no recorded instance in which Synura forms so prominent a part of the plankton of a body of water as it does of that of the Illinois River. It may be that a closer analysis than has yet been given the potamoplankton of other streams will reveal its prominence there also. It is present (Kofoid 90 '95) in the summer plankton of the Great Lakes at temperatures 15° to 20° below the summer majcimum of the Illinois River. Trachelomonas acuminata Schmarda.*—Average number, 1,094,- 615 ; silk, 873. This species appears in the plankton in April or May and continues into October or November. There is no vernal pulse, and the data are too irregular to trace the seasonal fluctuations. The greater numbers occur during the period of maximum heat. Excepting a single occurrence in February, this species has been found only above 40°, and its period of continuous appearance from May to October lies above 60°. It is evidently a summer plank- tont. Trachelomonas hispida Stein.*—Average number, 1,002,1 15 ; silk, 1,251. This is a perennial organism, found in every month of the year but in larger numbers during the warmer months. It was more abundant than usual in the winter of 1897-98 following the low water and unusual development of the previous fall. There are no large pulses in 1898, but in 1897 there is indication of a vernal max- imum on April 27 and an autumnal one of 85,500,000 on September 7. The data are too irregular to trace the seasonal fluctuations in detail. There is no doubt, however, from the evidence at hand that this is a predominantly warm-water planktont similar to the other members of the genus. Trachelomonas volvocina Ehrbg.*—Average number, 17,672,692; silk, 7,162. This is the most abundant species of the genus and is found throughout the year in almost every collection. It is most abundant from May to October, during the period of maximum heat. There are no well-defined vernal or autumnal pulses, but recurrent maxima during the summer are to be found in both 1897 and 1898. There are four such pulses in the former year,' and in the latter five, as follows: May 17 at 64° (14,400,000), June 21 at 77° (147,600,000), July 19 at 84° (86,400,000), August 9 at 83° (252,000,000), and October 4 at 71.5° (11,700,000). The periods of greatest growth thus lie above 60° and the optimum is near 80°. None of these pulses coincides with a volumetric maximum of the silk-net catches (Pt. I., PI. XII.). They usually follow these maxima at intervals of one or two weeks— a phenomenon often observed in other synthetic species. It may be explained by the decrease in animals w^hich feed upon the organ- isms in question. These volumetric pulses are predominantly 91 animal in their composition, and when they decline the organisms upon which the disappearing animals were feeding have an oppor- tunity to multiply with less decimation in their ranks. This species is one of the most abundant of the synthetic organ- isms in the summer plankton, and next to Euglena is the foremost among the synthetic elements of the food cycle of the plankton. The presence of many light-colored or even colorless forms (forma hyalina Kl.) justifies the suspicion that members of this genus, like those of its near relative Euglena, adopt holozoic nutrition in the presence of abundant organic matter suitable for food. This species, as well as the others above listed, is exceedingly variable in the proportions of the lorica, in its color, and in the development of the neck. It is very desirable that its life history and the full limits of its ^'ariation be determined before many more new species are proposed in the genus. In addition to the forms above listed, the following have been noted as present in small numbers in the summer plankton, viz. : T. armata Ehrbg., T. caudata Ehrbg., T. torta Stokes, T. iirceolata Stokes, and T. volvocina var. riigiilosa Kl. Uroglena americana Calkins.—-This species was found in small numbers in July and September, 1897, and in January, 1899. Uroglena radiata Calkins.—This species was found in January, 1896 ; in April and May, 1897 ; and in March and April, 1898. There was a vernal pulse of 15,279 on April 29, 1896. Uroglena volvox Ehrbg.—This species w^as found sparingly in the spring plankton in 1896. Uroglena is one of the few organisms ^^hich the usual method of plankton collection and preservation fails to keep in fair condition for subsequent indentification. The gelatinous matrix is easily crushed, and debris adheres to it so as to obscure it beyond recognition. Judging from the frequency of Uroglena in the living plankton it is very probable that the genus is much more abundantly represented in the Illinois River than the data at hand indicate. The genus seems to prefer the cooler waters of autumn and spring to those of midsummer. Volvox aureus Ehrbg.—This species was found from March to August, but in small numbers and irregularly. Volvox globator L.—This was somewhat more abundant than the previous species, and was found more frequently, especially during 92 1895 and 1896. It occurred from the first of May till the end of August, but always in small numbers. It is occasionally abundant in backwaters where there is much vegetation. In addition to the Mastigophora above listed there were many individuals belonging to unidentified species. They were as a rule the smaller forms, which are not readily identified in preserved material and under the conditions of plankton enumeration. They consti- tute about twenty-six per cent, of the total Mastigophora enu- merated. In silty planktons their number is relatively somewhat larger on account of the difficulties attending the determination of species in such material. These unidentified flagellates occur in every collection, and are somewhat more abundant in the summer months. [rhizopoda. Average number, 55,364, including filter-paper collections; 23,826 without them. This group of Protozoa is numerically of less importance than the ciliates or flagellates, but its quantitative significance is greater than the numbers of individuals indicate. This is due to the relatively large size of the Rhizopoda, and also to the fact that plankton collections afford only an irregular and in- complete record of the rhizopodan fauna of any body of water, and give but an imperfect idea of the part which these organisms play in the total economy of the lake or stream. This results from the fact that they are as a rule largely bottom or shore-loving species, and are generally either adventitious or temporary constituents of the plankton. The seasonal distribution of the total Rhizopoda in the Illinois River gives evidence of the adventitious or temporary nature of the contributions of the group to the plankton. There are pulses in 1898 on January 25 (66,388), February 22 (141,524), August 23 (36,800), September 27 (59,200), and November 15 (42,000), all of which appear on rising water and are largely adventitious, their presence in the plankton being due to the disturbances of currents, waves, and the like. There are pulses on May 10 (49,800), June 28 (37,000), and July 19 (28,800) which cannot be traced to any general hydrographic condition. These, as will be suggested in the discussion of the seasonal fluctuations of individual species, are probably due to the temporary adoption of a limnetic habit on the 93 part of some of the rhizopods,orto the appearance of Hmnetic forms, varieties, or species—according to the systematic value placed upon these eulimnetic individuals. I am inclined myself to regard them as seasonal forms of species which are predominantly of the bottom or littoral fauna, which have multiplied rapidly under the stimulus of abundant food. Owing to this fact, to the storage in their tissues of the products of metabolism, such as gas and oil vacuoles which tend to lighten their specific gravity, and to the frailer structure of their shells under conditions of rapid multiplication, they abandon their customary benthal or littoral habitat and assume temporarily a limnetic distribution in the plankton where they con- tinue to find abundant food. Their, appearance here under these circumstances is a result of their physiological condition, and with its cessation they decline, as shown by their pulse-like occurrences. Whatever the systematic valuation placed upon these limnetic forms may be, there is no doubt of their occurrence. They have appeared in every year of our operations, but were most prevalent in 1897, a year of most stable conditions, and also in the quieter backwaters, and on the declining spring flood orJune risewhen hydro- graphic conditions are less catastrophic than those of early flood stages. In 1897 there was a pulse of 68,400 (silk-net only) on August 8 and another of 1,268,400 on September 7, both in stable conditions and almost exclusively of limnetic types, diftering in this respect from the pulse of 141,524 on February 22, 1898, which was pre- dominantly of an adventitious character, resulting from the flood of that period (Pt. I., PI. XII.). The contrast in the numbers of Rhizopoda in the plankton during warm and cold seasons of the year is very striking in 1897. The average per m^ per collection from May 1 to October 1, that is, above 60°, is 161,045, omitting all filter-paper collections, while in the seven months of lower tempera- tures this average is only 4,771. During the warmer period the June rise was the only hydrographic disturbance (Pt. I., PI. XI.) to which any adventitious increase might be attributed. This con- trast is less evident in 1898, when the summer hydrograph was more disturbed. These larger numbers during warmer months may be attributed in part to the greater numbers of the Rhizopoda in their littoral habitat, and in part, doubtless, to the fact that at low water the shore and bottom fauna are brought into more intimate relation with the plankton, and in the river the disturbance of these regions 94 by current, waves, seines, boats, and fish make relatively larger contributions at low-water stages to the diversification of the plankton. In addition to these factors, however, there is abun- dant indication that many individuals assume during the warmer months a eulimnetic habit, and that some of the Rhizopoda become, for the time being at least, typical, though temporary, planktonts. It naturally follows that in so far as the plankton is concerned, the Rhizopoda exhibit a seasonal preference for the warmer months above 60°. Maximum numbers were attained only at the higher temperatures save in those instances where they attend w4nter floods. In a measure the seasonal distribution of the Rhizopoda in the plankton reflects that of the group in its normal habitat ; but at the best the picture is incomplete. The Rhizopoda have important relations in the economy of the plankton. They feed upon diatoms, desmids, the smaller algae, and even the chlorophyll-bearing Mastigophora such as Trachelomonas and Carteria. Their occurrences in the plankton do not exhibit any striking correlation with those of the groups named. The great pulse of September 7, 1897, for example (PI. II.), lies in a depression of the diatoms and coincides with pulses of Chlorophycecs and Mastigophora, and that of August 10 (68,400) exhibits a similar relation, the diatoms rising the following week as the Rhizopoda fall. In 1898 the pulse of Rhizopoda on June 28 of 37,000 (Table I.) culminates a fortnight after that of the diatoms and ChlorophycecB and a week after that of the Mastigophora. It thus is intercalated between the June and July pulses of these chlorophyll-bearing organisms (PI. II.). The Rhizopoda pulse of July 19 (28,800), on the other hand, occurs with the coincident pulses of the three groups named (PI. II.). The immediate diluent effect of flood waters upon the plankton combined with their tendency to increase the number of adventitious Rhizopoda results at times in the intercalation of their pulses with those of the chlorophyll-bearing organisms whose relative numbers are reduced by the dilution. The data evidently do not afford any adequate solution of the inter- calations of the Rhizopoda with other organisms. The Rhizopoda are very frequently found in the digestive tract of limnetic rotifers, but I have never noted the Entomostraca feed- ing upon them. They are important elements in the food of young 95 fish (Forbes, '80) such as the Catostoniidcu and some of the Siln- ridcE and minnows. I have found them in great abundance in the intestine of the aduk gizzard-shad (Dorosoma) , and in the contents of the digestive tract of the German carp {Cypnnus carpio). In the pages which follow, the seasonal distribution, or occur- rence in the plankton, of thirty-one Rhizopoda is discussed, and the presence in the plankton of the Illinois of twenty-eight other rhizopodan forms which have been recognized by other writers as of specific rank is noted. This b}' no means exhausts the rhizopo- dan fauna of the environment which was the field of this investiga- tion. A continued study of the plankton itself would doubtless greatly extend the list of adventitious forms from the shore and bottom, and a more careful analysis of the variants, especially in the Difflugia globulosa-lobostoma group, would still further increase the richness of the fauna from the systematic point of view. Hem- pel ('99) lists sixteen species from this locality, and Penard ('02), in discussion, remarks: "Une pareille pauvrete dans une region riche en organismes de toute nature, est une impossibilitie mate- rielle." However, neither Hempel's paper nor the present one pretends to give a full account of all the Rhizopoda of the region. He dealt largely with plankton collections, and the present paper deals with them exclusively. There is but little in plankton literature which gives with any fulness the seasonal distribution of the Rhizopoda , or indicates that they are of any considerable importance in the economy of the plankton. The importance which they acquired in the plankton of the Illinois is no doubt in part due to the nature of the environ- ment with which we are dealing. The somewhat sporadic and meteoric character of their appearances in our waters leads to the inference that full seasonal analyses of the plankton of other bodies of water at brief intervals may reveal a greater prevalence of the Rhizopoda in the plankton than has hitherto been detected. DISCUSSION OF SPECIES OF RHIZOPODA. Amceha Umax Duj.—This was frequently abundant in the water- bloom of midsummer, but was not identified in the plankton collections. Amoeba proteus Rosel.—Average number, 342. The individuals here assigned to A. proteus include those taken in our plank- 96 ton which belong to the type of A. radiosa Ehrbg., a type which presents no distinctions sufficiently well-defined to separate it spe- cifically from the first-named form. It seems probable that A. radiosa includes small individuals of A. proteus w^hich are not, at the time of observation, creeping upon a substratum ; that is, they are limnetic, floating free with filamentous pseudopodia character- istic of that condition. Verworn ('97) has shown that A. proteus takes the radiosa form in weakly alkaline solutions. " Pond water rich in algae may have an alkaline reaction (Knauthe, '98) in bright sunlight. Larger individuals, distinctly referable to the A. proteus type w^hen taken in the plankton, possess at times the slender pseu- dopodia of the A. radiosa type as well as the blunter ones charac- teristic of the A. proteus form. I see no valid reason for separating the two as distinct species. Most of the Amoeba recorded from the plankton collections belong to the A. proteus type, the smaller ones belonging to the radiosa type probably escaping through the meshes of the silk net. This species was found in 30 of the 180 collections examined, being observed in all months of the year except May, November, and December. The conditions attending its occurrence suggest that it is not, habitually at least, an active planktont at all seasons of its occurrence, but rather a tycholimnetic member, an invader from the littoral or bottom fauna, or a temporary accession during the w^armer months. In the first place, both the number of occur- rences and the numbers of individuals found are small, and the seasonal distribution, plotted from the data of the collections of the five years, is exceedingly irregular. Furthermore, 17 of the 30 occurrences happened on rising floods, w^hen the fauna of the bot- tom and shore of both the river and its tributaries is most mingled with the plankton. Further evidence of the agency of floods in introducing Amoeba into the plankton is brought to light by a com- parison of its occurrences in 1897 and 1898. As shown by Plates XL and XII., Part I., the hydrograph of 1897 is much less irregular than that of 1898, the latter year exhibiting repeated fluctuations in level due to floods. As a result we find Amoeba occurring rela- tively (to the number of collections) almost twice as often in 1898 as it did in 1897. It may also be significant that Amoeba was not found in November and December, months of unusual stability in river levels. There is, however, a suggestion in the data of distri- 97 bution (see Table I.) that Amoeba may become an active member of the plankton during the warmer seasons, like other Rhizopoda, as a result, perhaps, of the formation of gas or oil vacuoles in its proto- plasm. Of the 30 occurrences, 21 fall between April 19 and Octo- ber 17, with water temperatures of 58° and 56°, respectively. Of these 21 occurrences in warm waters but 8 accompany flood inva- sions, while all of the 9 occurrences during the colder months are in connection with such disturbances. Finally, the maximum num- ber per cubic meter (6,400) was found July 21 in clear waters, free from the debris of flood invasion. In conclusion, it seems probable that^4;n^6a in warmer seasons of the year (above 56°) may adopt a limnetic habit. There is, however, the possibility' that local and minor disturbances of the water due to current, waves, etc., are the occasion of its presence in the plankton in the absence of flood conditions. Jennings ('00a) reports both .4. proteus and A. radiosa in the open water of Lake Erie. The range of temperature of river water in which Amoeha was found was from ?>2° to 89°^the full extremes observed by us in the river at Havana. The temperature at the maximum occur- rence, July 21, 1897, was 82°. It is perhaps significant that 14 of the 30- occurrences of Amoeha were between June 21 and Sep- tember 6, the period of maximum heat, the river averaging aknost 80°—apparently the optimum temperature for the occur- rence of Amoeha in the plankton in this locality. The relative numbers of individuals found in the various collections of the five years are too irregular to suggest any conclusions as to a seasonal C3^cle. Amoeba verrucosa Ehrbg.—Average number, 19. This species was found but three times in the plankton, once each in May, August, and September, occurring but singly, and in each case in flood waters. It is apparently a tycholimnetic member of the plank- ton. The temperature limits of its recorded occurrence in the plankton were 58° and 82° respectively. Arcella. This genus is represented in the plankton by four species and two varieties which, like most of the Rhizopoda, are exceedingly variable, grading in some instances into each other by occasional 98 individuals which present intermediate characters. The majority of the individuals were taken in a living condition, though many empty shells were found. The conditions of the examination of the plankton and the opacity of many of the shells made it impossible to distinguish the dead shells in all cases. The records include many dead shells. Arcella costata Ehrbg.—Average number, 48. For the purposes of this paper I have included here all those individuals which possess an angular or ribbed shell. Leidy ('79) refers such forms to A. vulgaris. Individuals of this type are rare, occurring infrequently and in small numbers. It was recorded but 18 times in the 180 collections, and the largest number per cubic meter was only 1,187. As in the other species of the genus, the warmer months are favored, fourteen occurrences falling in June-September in water at 70° or above. The other four records are one each in April, October, November, and December. The seasonal range of this form in the plankton thus falls in the main within the period of the maximum abundance of A. vulgaris, of which species it may be but a variant. Arcella discoides Ehrbg.—Average number, 972. This prevalent species is not in all instances easily separated from A. vulgaris. Indeed, even Leidy ('79) states that it graduates into A. vulgaris, and that he views it as the variety of this species in "which the shell presents a greater proportionate reduction in height compared with the breadth." In the enumeration of our plankton catches, the larger, flatter, and unornamented individuals have been referred to this species. Both the brownish and the hyaline forms should probably, for reasons hereafter given, be included here, and they are so grouped in the present discussion. Thus considered, A. discoides is the most abundant member of the river plankton be- longing to this genus, including two thirds of all the individuals observed. This species occurred in almost two thirds of the collections, hav- ing been recorded in 115 of the 180, and more frequently and in larger numbers in the latter half of the five years than it was in the earlier period. This is in part explained by the unusual fluctua- tions of the river levels in 1898, during the maximum summer occurrence of the species. Like the other species of the genus, A. discoides has a period of maximum occurrence in the latter part of summer, as is shown in Table I. Of the 115 occurrences, 55 were in 99 June-September, in water at or above 70°, while in the remaining eieht months there were but 60 occurrences. This contrast is heightened by the ratio of occurrences to the total number of collec- tions, \\hich in the period from June to September inclusive is 55 to 68 and in the remainder of the year only 60 to 112. The num- ber per cubic meter is also higher during this warm period, averag- ing for a single occurrence 1,376 to 1,028 for one in the remainder of the year. The average for the colder months falls to 850 if the large accessions attending the floods of February and November are omitted in the totals. The same causes efficient in determining the summer maximum in other Rhizopoda of the plankton are doubtless operative here, and as in A. vulgaris the impetus of the summer in- crease is carried over into the autumn, causing a slight increase in numbers as compared with the numbers at corresponding temper- atures in the spring months. It seems probable that high temper- atures favor its occurrence in the plankton, not, however, directly, but because of greater abundance of food under those conditions, greater metabolism, and the storage of the products as oil or gas vacuoles which tend to lower the specific gravity and thus to bring the animal into the plankton. The adventitious occurrence of A. discoides in the plankton is shown by the fact that 45 of the 115 occurrences are with rising flood waters. The greater part of them lie in the colder months; in fact, nine tenths of the occurrences between October and May are correlated with flood movements. For reasons above given, how- ever, A. discoides may be regarded as temporarily adopting a lim- netic habit during warm months as a result of its physiological condition ; at least many individuals of the species exhibit this habit during the warmer months. The data do not indicate that the open water is at any time the center of distribution of the species. There are no indications of recurrent pulses in the species and, as might be expected in case of adventitious planktonts, but little evidence of a characteristic seasonal distribution. There is some evidence that the summer is the period of most active multiplica- tion, and that an exceedingly transparent and hyaline form other- wise resembling A. discoides is the young of this species. In 1898 separate records were kept of the two types with the result that they were about equally abundant—24,159 and 26,387 for the brown and hyaline types respectively. (8) ¥. 100 With but few exceptions the seasonal distribution exhibited by the hyaline form was very similar in time and numbers to that of the brown form. Both occurred more frequently and in larger numbers in the warmer months, and irregularly and in small numbers in the colder waters. Both entered in larger numbers with flood waters. The differences though sHght are suggestive. The hyaline form was less frequent than the brown both in occurrences and numbers dur- ing cold weather, and summer floods sometimes brought a rela- tively larger number of the hyaline type. These are conditions that might be expected if the latter is only the young (that is, the daughter organism occupying the new shell after fission of the oc- cupant of the old) of Arcella discoides. In warmer months food is more abundant and, presumably, fission more frequent. For this reason the young individuals abound at that time. Owing to the difference in the specific gravity of the two, the hyaline type is more readily transported by flood waters. Though not con- clusive, the data here presented seem to favor the view that the hyaline form is only a stage in the life history of the individual Arcella discoides. The species A. artocrea Leidy and A. polypora Penard occur also in our waters, but were included with A. discoides in the enumera- tion. Typical representatives of these species are not, however, present in any numbers Arcella mitrata Leidy was found but once—on Aug. 1,1895, in small numbers, at 78.5°. Arcella stellata Perty.—Under this designation are included only those individuals which have well-defined prolongations on the margin of the shell. Only a single occurrence in small numbers (48 per cubic meter) was recorded for the typical A. stellata— July 29, 1895, at a temperature of 75.5°. Arcella vulgaris Ehrbg.—Average number, 1,098. This species is somewhat more abundant than A. discoides, but occurred in fewer collections. It is a somewhat common planktont, whose seasonal distribution exhibits some irregularities attributable in part, as in the case of other members of the genus, to flood conditions. It was found in 61 of the 180 collections examined, and in approximately one third of those made in each year, excepting in 1894, when itj was not recorded, and in 1898, in which year it was found in about! 101 half the collections, the river levels for this latter year being subject to more than the usual disturbance. Arcclla vulgaris is found throughout the whole year, with a marked predominance of occurrences during the warmer months, June to September inclusive, for during this period, in which a total of 68 collections w^ere made, this species was found in the plankton 34 times. If the month of October be included, the ratio is 44 oc- currences in 83 collections, while in the remaining 97 collections, from November to June, only 17 occurrences were recorded. Of the 10 occurrences in October, 7 were in water at or above 55°. The season of frequency in the plankton thus ranges from June through October. In both frequency of occurrence and in numbers of individuals (see Table I.) there is an apparent maximum in August, preceded by an increase in June and July and followed by a decline in September and October. Arcella vulgaris thus seems to be a late summer planktont. The continuance into October may in part be due to the temperature conditions above cited, and perhaps also to constant seining of the river by fishermen in the low-water stages at that time, causing repeated disturbances of the bottom and shores, where Arcella habitually lives. This maximum frequency of Arcella during the warmer months in the plankton is, however, probably due to the formation of gas or oil vacuoles in the plasma under the conditions of higher temperatures. Their flotation is thus facilitated, and they become, in a way, semi-active but temporary planktonts. That floods are also in part responsible for the presence of Arcella in the plankton is evident from the fact that 32 of the 61 occurrences come with rapidly rising waters, or shortly after rapid rises, during the interval of rapid decline. The larger numbers of individuals also appear in flood-waters, occurrences of more than 1000 per cubic meter happening 10 times with floods to only 4 in more stable conditions. The maximum occurrence, 25,272 per cubic meter, came with the flood of February, 1898, indicating the presence of this species in large numbers, even under winter conditions, in some local environment tributary to the flood plankton. The average number per cubic m.eter in the 61 collections con- taining ^rr^//a was 1,260; and the maximum, 25,272, as above noted. This species occurred in only 10 collections in stable conditions of the river, when the temperature of the water was below 55°. The 102 average number oi individuals in these cases was, however, only 230 per cubic meter as against 1,443 when the temperature was above 55°, or, if below, w^hen floods prevailed. The seasonal and numerical distribution of occurrences and individuals alike point to the agency of floods and higher temperatures in the introduction of Arcella into the plankton from its usual habitat, the bottom and the shore. This species occurred in water ranging in temperature from 32° to 89°. Being a bottom form, the plankton data do not afford a satisfactory basis for determining its true seasonal distribution and optimum temperature. The maximum number found, 25,272, was in water at 32°; but this was an isolated occurrence in a flood, and serves only to illustrate the irregularity of distribution in the plankton of tychohm netic organisms. Centropyxis aculeaca Stein.—Average number, 570. This species has appeared in collections in every month of the year, but its sequence is frequently mterrupted and its numbers are quite irregu- lar. Practically without exception all the larger occurrences attend rising flood waters. It is evidently adventitious at all seasons of the year. Centropyxis acuLeata var. ecornis (Ehrbg.) Leidy.—Average number, 604. In foiTner years this species was less frequent than the preceding species. Its appearances in the plankton tend to coincide with those of C. aculeata (Table I.), and are doubtless due to the same causes. Thus in the February flood of 1898 there is a pulse of 12,636 of C. aculeata and one of 9,477 of var. ecornis. C. IcBvigata Penard seems to be identical with this variety. The data concerning both C. aculeata and its variety ecornis are too irregular to throw any light on the seasonal cycle of these adventi- tious planktonts. Cochliopodium hilimhosum (Auerbach) Leidy.—Average number, 1,384. This species was found in the plankton during 1898 in irregular numbers in 27 of the 52 collections. The distribution of the occurrences affoids indubitable proof of their close dependence upon flood waters. Fn 15 of the 27 cases Cochliopodium appeared with a rising river, and in all but 6 cases, in periods of considerable movement in river levels (cf. Table I. with PI. XII., Pt. I.), such as the rising flood of January and February and the repeated minor 103 fluctuations of August and the following months. The year 1898 was one of unusual irregularity in the hydrograph (Pt. I., PI. XII.), especially at the lower stages of the river, at which times this rhizopod appeared most frequently. Its maximum occurrence, 20,898 per cubic meter on Jan. 25, accompanied a rise of 0.6 of a foot in 24 hours. At other times the numbers range from 100 to 8,000 per cubic meter, their irregularity affording additional ground for regarding this species as an adventitious planktont. Cochliopodiiun was present in water ranging from 32.1° to 89°, the maximum number observed being found in water almost at the freezing point, when the river was full of running ice. That this is the optimum temperature for this organism is not, however, to be inferred, since, as has been shown above, this species is adventi- tious in the plankton. Plankton collections do not afford adequate data for determining the seasonal cycle of the organisms habitually li\'ing upon the bottom. This species was not found, though careful search was made for it, in the winter collections of 1899. Its absence from the records of years previous to 1898 may in part be due to a failure to observe it in the silt-polluted collections in which it is most apt to occur. Cyphoderia margaritacea Ehrbg.—Average number, 198. This species has occurred in every month but February. In 1898, the majority of the occurrences and three fourths of the numbers ap- peared between May 1 and October 1 at temperatures above 60°. It was never abundant at any time, though there is this indication of its increased numbers during the warmer season. It is not an im- portant element in our plankton. Apstein ('96) found it somewhat irregularly in the plankton of German lakes. In our waters it exhibits no marked dependency upon floods for its presence in the plankton, though it is probably capable of assuming the limnetic habit in the warmer season. Cyphoderia trochiis Penard appeared occasionally with the pre- ceding form, from which it is distinguished by its conical horn on the fundus and by its larger scales. Di-fflugia. This genus is the most abundant one of the Rhizopoda in the plankton of the Illinois River, and is a factor of quantitative 104 importance in its economy. It includes a number of forms notorious for their variability and for the difficulty with which specific dis- tinctions can be applied. I shall discuss the species as they were enumerated, and shall correlate my work with Penard's ('02) recent elaborate analysis of the species so far as I can with the aid of my notes in the absence of the collections. Opinion as to the validity of the species is expressly withheld excepting in those instances in which it is formally stated. Difflugia acuminata Ehrbg.—Average number, 315. This spe- cies has occurred in every month of the year and in 83 out of 180 collections. In 1898, two thirds of the occurrences and three fourths of the individuals were taken between May 1 and October 30, at temperatures above 70°. In this year there are six recurrent pulses from June to November, but all but one of these are found on rapidly rising flood waters, and they bear no constant relation to the pulses of diatoms previously noted, with which in some instances they are intercalated, though this is not regular or constant. Similar tendencies to appear with floods and in greater numbers and more frequently in summer can be detected in records of other years. It was more than twice as abundant in 1896—a year of interrupted hydrograph (Pt. I., PI. X.)—as in 1898. This is one of the larger and heavier rhizopods, and its occurrence in the plankton is doubt- less adventitious, due to floods and currents, and its greater numbers and frequency in the summer may result from its greater abundance at that season in its natural habitat, the shore and bottom, and perhaps, also, from its lighter specific gravity during the warmer season. An illustration of this appears on the rising flood of June, 1897, when the maximum number recorded (10,000 per m.^) oc- curred. The shell of this species is exceedingly variable in size, constitu- ent particles, and proportions. A number of forms separated by Penard ('02) and others as -distinct species were grouped under D. acuminata in the enumeration. The greater number of these belong to the type designated by this name by Penard ('02). D. acuminata var. inflata Penard and the somewhat similar D.elegans Penard are not uncommon. D. acuminata var. umhilicata Penard, D. elegans var. teres Penard, D. ctirvicaulis Penard, D. lanceolata Penard, and D. scalpellum Penard occur also, but are rare. 105 Difflugia hicuspidata Rhumbler.—Average number, 76. A sep- arate record was kept of this bicuspid type in the later years of our collections. Penard ('02) regards it as a synonym of his D. elegans, though it would seem to be as worthy of specific distinction as many other variants to which he accords this rank. It varies greatly in the relative development of the accessory "horn," which is sometimes but a mere elevation near the base of the main horn. Individuals with equal and symmetrical horns represent the other extreme. In a few cases tricuspid individuals have been seen, evidencing a tendency to vary towards the type found in D. varians Penard and D. fragosa Hempel. This form was about one fourth as abundant as D. acnminaia, and eight of the ten occurrences fall between May and October, usu- ally with D. acuminata and presumably for the same reasons. Difflugia constricta Ehrbg.—Average number, 46. This species occurs irregularly at all seasons of the year .without marked prefer- ence for the warmer months, and often, but not always, with flood waters. It occurs throughout the whole range of temperatures, and the largest number (2,778 per m.^) appeared during the decline of the spring flood. Data are too infrequent to establish any seasonal routine. This species varies greatly, and is connected by an unbroken series of variants with the genus Centrupyxis. Penard ('02) also notes the existence of this connection, and states that after careful search he was unable to find any constant distinction which would suffice for its separation. In my enumeration only the elongated and smooth individuals were referred to this species. The spinose forms were referred to Centropyxis aculeata, and those similar in form to the spinose type; but those free from spines, to C. aculeata var. ecornis. Difflugia corona Wallich.—Average number, 36. -In 1896, when the hydrograph was much disturbed, the average number was more than twice as great. This superb species was found in every month of the year except December, Ivat never in large numbers. Its large size (200-300 «), and its heavy shell militate against its pres- ence in the plankton, and its occurrences are irregular and its num- bers few. There is no marked preference for warmer months, and four fifths of its occurrences are in rising flood waters. It is plainly 106 an ad\-entitious planktont. The data are too irregular to trace its seasonal distribution. As a species it is as well defined as any in the genus. It is not in our waters connected by intermediate forms with other species. Its assignment to D. lobostoma by Schew^akoff ('93) is not in my opinion justifiable unless we regard all forms of Difflugia as belonging to one species. Difflugia fragosa Hempel.—Average number, 25 ; in 1896 over 100. This species occurred in every month of the year but Febru- ary, though three fifths of the records and the majority of the in- dividuals were found between May and October at temperatures above 60°. The data are too irregular to trace the seasonal history of the organism, but they suffice to suggest the agenc}^ of floods at all times and of high temperatures during the summer, as factors in the occurrence of the species in the plankton. The shell of this form is relatively to that of other species rather heavy, and this fact combined with the irregularity of its occurrence seems to justify the conclusion that it is largely adventitious at all seasons of the year. The species exhibits a great deal of variation in the development of the central spine—Hempel ('99, Fig. 1)—and in the number and arrangement of spines in the accessory circlet. The mammillate form of the central spine figured by Hempel is not usually present. Individuals in which the central spine is but feebly developed seem to connect this species with D. varians, recently described by Penard ('02). Otherwise, and in our waters, the species is well delimited. Difflugia glohiilosa Duj.—Average number. 7,194; in 1897, 47,329, the larger number in this year being in part due to a remark- able pulse of 1,240,000 early in September. This is the most abundant of all the rhizopods in our plankton, occurring most frequently and in largest numbers. It is found in every month of the year, and in 1898 appeared in every collection except four in De- cember. With a few exceptions in the autumn of 1898 (Table I.), no large development (exceeding 10,000 per m.^) has taken place earlier than May or later than September—that is, at temperatures below 60°. The occurrences are most continuous and the numbers of individuals are largest during the warmer period between the months named. The largest pulse, that of 1,240,000 on September 107 7, 1897, was at 80°. A pulse of 48,000 on November 22 at 40° gives evidence of considerable range in adaptation to temperatures. In Table I. the seasonal distribution of D. globulosa is given in full. It differs from that of previous years mainly in the fact that the summer pulses do not here have the amplitude reached in other years; for example, in 1896 (252,000) and 1897 (1,240,000). It is characterized by considerable irregularity caused by somewhat abrupt pulses at irregular intervals. A comparison of these occur- rences with the hydrographic conditions (Pt. I., PI. XII.) indicates that in the colder months increase in numbers in the plankton at- tends flood waters only, as, for example, in January, February, late October, and November. In the summer, pulses may also come with floods. For example, that of 252,000 on May 25, 1896, ap- peared on the upward slope of the June rise of the year, and that of 80,000 on June 28, 1897, came with the belated June rise of that year. On the other hand, some of the minor fluctuations appear on declining floods, and the maximum one of our records, that of Sept. 7, 1897, came in the midst of the most prolonged period of stable low water (Pt. I., PI. XI.) found in the six years of our operations. From these facts it is evident that floods are efficient in increasing the number of D. globulosa in the plankton, and that the amplitude of the pulses to which they contribute is much greater in the warmer months (above 60°) than in the colder ones—as a result, perhaps, of the greater numbers present in their normal habitat, the shores and bottom, and also as a result of their readier flotation at this season. In so far as their presence is due to floods they are adventitious. On the other hand, it is very probable that they become temporarily eulimnetic in habit during the summer months. The evidence for this lies in their greater numbers in a period which is predominantly one of greater stability. Thus in 1898, in the 22 collections between May 1 and October 1, the average number present is 9,731, while in the remaining seven months of colder weather the number is only 5 ,200. Additional evidence arises from the fact that pulses of unusual magnitude have occurred quite independently of any factor such as flood or other disturbance which might cause their adventitious introduction into the plankton. Thus on Sept. 7, 1897, there is a symmetrical pulse whose rise and decline occupy four weeks, as shown in the following table. The total change in river levels in this period of four weeks (Pt. I., PI. 108 Date 109 The question of specific limits and variation in this organism is one of exceeding difficulty, and I see no satisfactory solution for it until some one attacks the problem by a study of the variation by modern quantitative methods, and endeavors by breeding under control to establish the limits of variation within the normal ranse of seasonal changes of the environment. When this is done, some more satisfactory criterion for species in this group of planktonts will be feasible than the present condition affords, in which slight difterences from previous descriptions are held to be valid for specific distinctions. Thus, in recent years, species of plankton Difflugia have been described by Heuscher ('85) {D. urceolata var. helvetica) from Swiss lakes ; by Zacharias ('97) {D. hydrostatica) from Lake Plon ; by Garbini ('98) {D. cyclotellina) from Italian lakes; by Levander ('00) {D. lohostoma var. limnetica) from Finnish waters; and by Min- kiewitsch ('98) {D. planktonica) from Russian waters. All of these forms occur in the Illinois River, and there are others equally worthy of specific designation in our plankton as yet undescribed. They occur most abundantly at the times of the pulses, especially of those in stable conditions. In my opinion they are all mere limnetic varieties of D. globulosa or D. lohostoma, the form of the shell and its constituent particles being modified by the habit of life in which these individuals of the seasonal cycle are found. They occur at times of abundant food, rapid multiplication, and limnetic environ- ment. Their shells are accordingly lighter, more chitinous and transparent, and the foreign particles adherent to them partake of the nature of those of the silt in suspension. This, however, is merely an opinion based upon an examination of the statistics of occurrences, and upon the work of plankton enumeration in which all individuals must be assigned to some species. This is at least a different point of view from that of the systematist, who may, per- haps, lay more stress upon divergences from described types and less upon links connecting such variants. For the sake of genuine progress in the science it would seem to the wTiter extremely desir- able that more attention be given to the question of variation and less to the description of new species under criteria now in vogue. It may be desirable, indeed necessary, to distinguish such forms in the plankton. It would be both safe and conservative to designate them as forms, or, at the most, as varieties. no The location -of the pulses of D. globulosa bears no constant rela- tion to those of other organisms, owing, in part, at least, to the irregularities of the floods upon which some of them seem to depend. The great pulse of Sept. 7, 1897, is intercalated between two pulses of diatoms and other chlorophyll-bearing organisms, and some others bear a similar relation to their food supply, while some co- incide with an increase in these synthetic organisms (cf. Table I. and PI. II.). Difflugia globulosa and the following species were reported by Smith ('94) in the plankton of Lake St. Clair; by Jennings ('00a) in that of Lake Erie; and were common in the plankton of Lake Michigan (Kofoid '95). Difflugia of the forms included here under D. globulosa and D. lobostoma have been reported by many authors from various European lakes and rivers, but in no reported instance do they reach the numbers or importance in the plankton that they do in the Illinois. Full records of their seasonal distribution may, however, bring such importance to light. Difflugia lobostoma Leidy.—Average number, 1,158. In the total of all collections it is about one fifth as abundant as D. globu- losa. Like that species it occurs throughout the whole year in almost every collection (Table I.), and the fluctuations in its occur- rence follow very closely those just described for D. globulosa in the direction of their movement. The amplitude of the pulses is less, as a rule, and their culminations and limits are coincident, or at least approximate. Thus, on Sept. 7, 1897, D. lobostoma attains only 24,000, and the pulse of D. globulosa on June 28 (80,000) is attended by one of 96,000 in D. lobostoma in the next collection, on July 14. There are in this species also the same influx into the plankton with floods, and increase in numbers at temperatures above 60°. There are 954 per collection per cubic meter below this temperature to 1,436 during the warmer months in 1898. There are also pulses during the warmer months, in stable conditions, coincident with those of D. globulosa. Similar causes presumably contribute to these results in both species. - Difflugia lobostoma is also exceedingly variable in proportions, in the texture of the shell and the degree of incision, and in the num- ber of lobes about the mouth. Two, three, and even four have been noted, and they vary greatly in depth, in regularity, in perfection of their development, and in the structural border which sometimes Ill forms their margin. Chitinous, brownish, or more or less trans- parent shells are abundant when pulses occur. Forms which connect this species with D. globtilosa have been observed. In- cluded with D. lohostoma are forms which have since been described by Penard ('02) as D. gramen, D. gramen var. achlora, and D. lithoplites, though I have not found in the Illinois plankton any of the last-named with the peculiar tipped horns found by Penard upon many individuals of his species. Diffliigia pristis Penard (?).—A small Difflugia was found occa- sionally in the filter-paper collections in the colder months, but only from November to March. It was often dark, or even blackish, resembling in this respect Penard's D. pristis. Individuals not thus darkened approach more nearly D. fallax Penard and L*. pulex Penard. Difflngia pyriformis Perty.—Average number, 368. This species occurred in e\^ery month except January, but generally in small numbers and irregularly. The largest number taken—12,000, on May 25, 1896—came with the flood at that time (Pt. I., PI. X.), and all the large occurrences of 1898 came with rapidly rising water (cf. Table I. and Pt. I., PI. XII.). There are no indications of pulses during stable conditions, and we must conclude that the species is purely adventitious in our plankton. It is one of the largest species with a heavy shell, and its flotation is impeded thereby. This species is exceedingly variable. The following varieties or variants, given specific rank by some writers, have been noted, and are included with D. pyriformis in the enumeration: D. pyriformis var. nodosa Leidy, D. pyriformAs var. daviformis Penard, D. pyriformis var. venusta Penard, and D. pyriformis var. lacustris Penard. A more slender and smoothly contoured form than the last is not uncommon. D. capreolata Penard and D. bacillifera Penard were also found, but are rare. Diffliigia rubescens Penard was taken but once—on May 25, 1896. Difflugia tuberculosa Hempel was also found but once in the planktons enumerated, though Hempel ('99) reports it as appearing occasionally from August to November in 1895. Difflugia urceolata Carter was taken only in April and May, 1896, in small numbers at temperatures of 66°- 80°. 112 Dinamoeha mirahilis Leidy was found in the plankton but once—Apr. 12, 1898, in small numbers, at 52°. Euglypha alveolata Duj. was found in small numbers in the plankton, but only on Nov. 1, 1898, and March 14, 1899, at tempera- tures of 45° and 36°. E^iglypha ciliata Ehrbg. appeared in the filter-paper collections in 1897, in July, August, and November, in small numbers at tem- peratures ranging from 80° to 48°. This is said by Penard ('02) to be predominantly a sphagnum species, but widely distributed elsewhere in small numbers. Euglypha IcEvis Perty.—This minute rhizopod was found in the filter-paper collection of Oct. 4, 1898, at 72°. Nebela collaris Leidy was found only once—on June 25, 1898, at 32°. ;, ; Pontigulasia incisa Rhumbler.—This curious rhizopod occurred in the plankton in July and August, 1895, and again in August and September, 1897, at temperatures of 75°- 85°. Both occurrences were in stable conditions, and the temporary adoption of the lim- netic habit is suggested by their appearance at these times. Two other records in 1897—on March 22 and November 9, at 44° and 50°— extend the seasonal range of the species. These occurrences attended rising water and were apparently adventitious. Trinema enchelys (Ehrbg.) Leidy.—Average number, 158. This little cosmopolite rhizopod of the sphagnum fauna was found but eight times in the plankton. The individuals observed were all dark- ened by the granular food vacuoles to such a degree that structural details were obscured. It was noted only in the somewhat turbu- lent years of 1898 and 1899, though on account of its small size and the obscurity of its structure it may have been overlooked in previ- ous collections. The few occurrences are insufficient to establish any seasonal routine. They were at both extremes of the tempera- ture range and in all seasons but spring, with a predominance in late summer and fall. The species is evidently adventitious in the plankton, as shown by irregular distribution and small numbers, and by the fact that its occurrences coincide in all instances but one with rising water. 113 HELIOZOA. The Heliozoa of the plankton of the Illinois are few both in number of species and of individuals. They apparently play but a small part in the economy of the plankton. The average number for 1898 was but 4,883. Their occurrences are confined in the main to midsummer and early autumn. But four species were identified, though several others remain undetermined for lack of sufficient material, especially of the living forms. Apstein ('96) reports Heliozoa in considerable numbers in German lakes, with maxima in July-August. It is probable that these delicate forms are fre- quently crushed in manipulation or hidden in silt in our collections. DISCUSSION OF SPECIES OF HELIOZOA. Actinophrys sol Ehrbg.—Average number, 62. This species occurred irregularly from April to the early part of November at temperatures above 46°. It was recorded most frequently in the latter part of the summer, the largest number (28,000) appearing Sept. 7, 1897, at 80°. Actinosphcsriiim eichhornii (Ehrbg.) Stein.—Recorded a few times, from July to October, at maximum temperatures (75°- 80°), but always in small numbers. Endophrys rotatoriormn Przesm.—This heliozoan (?) has been recently described by Przesmycki ( '01) as parasitic, during a part of its existence, in Philodina and Hydatina. A parasite resembling this parasitic stage of Endophrys was observed by me in a bdelloid rotifer (Rotifer tardus) on several occasions, but it was never abun- dant, nor' was its connection with any free-swimming condition noted. The heliozoan affinities of this organism seem very ques- tionable. Nuclearia delicatula Cienk.—Average number, 4,760. This species in 1898 appeared first on June 21, attained a pulse of 78,400 on August 9 at 82° and another abrupt one of 65,600 on September 27 at 73°, and made its last appearance October 25 at 48°. Occur- rences in previous years are confined to midsummer. Its optimum conditions of temperature obviously lie near the summer maximum, and its lower limits near 50°. Its appearance in the plankton is not traceable to flood conditions, and it is apparently eulimnetic in our waters. 114 Hempel ( '99) reports Raphidiophrys pallida Ehrbg. and R. elegans Hertwig and Less, in the plankton of Quiver Lake adjoining the river, and I have found an undetermined species of Acanthocystis and a small heliozoan resembling Nudearia in the river plankton. SPOROZOA. Triactinomyxon sp.—In the plankton collections of each year there have been found free limnetic spores which unquestionably belong to that highly aberrant and peculiar group of organisms described by Stole ('99) as Actinomyxidia and regarded by him as Mesozoa, but later referred by Mrazek ('00) Caullery and Mesnil ('04), and Leger ('04) to X\iq Myxosporidia. The organisms de- scribed by Stole were parasitic in fresh-water oligochaetes, and it is not improbable that the limnetic spores taken in our plankton collections are derived from parasites in some of the numerous aquatic oligochaetes, or other invertebrates, found along the bottom and shores of the stream. The species here referred to Triactinomyxon differs in some details from T. ignotum Stole. It w^as found in the course of the six years at least once in every month of the year, but most regularly in May-September, and rarely and in small numbers in the colder months. Its transparency and long, slender, radiating, tripod-like arms give it a typically limnetic habit. Actinomyxidia, gen. et sp. indet.—Clusters of eight, or less, cylindrical spores radiating from a common center and bearing a marked resemblance in structural features to those of Triactinomyx- on, but lacking any anchor-like projections, were found sparingly in the plankton in June-September. The distinctively limnetic habit of these spore stages in the life- history of these parasites is unique among the Sporozoa, and has not, to my knowledge, been before noted. Many of the rotifers of the summer plankton, especially Brachi- onus and an occasional Asplanchna, have been heavily parasitized internally by small sac-like bodies, often pear-shaped, with the smaller end attached to the lorica, or of spherical or flattened form. They occur in such numbers at times as to be a menace to the rotifer population. They are usually most abundant in any given species at the time of, or subsequent to, its maximum occurrence. It 115 was not unusual to find as high as ten or fifteen per cent, of the indi\'iduals parasitized, and a number of empty loricae bearing addi- tional testimony to their destructive agency. Bertram ('92) describes these structures as " parasitische Schlauche" in the body cavity of rotifers, and Przesmycki ('01) works out their life history, and describes the organisms as DimcB- riiim hyalinum, but does not designate their systematic position or affinities. There are, however, marked suggestions of sporozoan affinities in the organism found in the rotifers of the Illinois plankton, which seems to be identical with that described by Przesmycki ('01). Obviously it is difficult to take a census of such internal para- sites. A record was kept, however, of the number of parasitized individuals in each species of rotifer, and references will be made to these results in the discussion of the hosts. Dimcerium appeared in both summer and winter rotifers, and its seasonal distribution naturally depends upon the number of available hosts. It was in consequence most abundant during the midsummer and autumn months. CILIATA. Average number, 15,812,346, including filter-paper collections. If these be excluded and the silk catches only averaged, the number will fall to less than a tenth of this sum. The ciliates are found in the plankton of the Illinois throughout the whole year, and as a whole they do not exhibit any common seasonal predominance. The analysis of the distribution of the individual species which follows, exhibits two diverse tendencies which afi'ect the distribution of the totals. These are the vernal and autumnal pulses of the TintinnidcE, represented by Codonella cratera and Tintinnidium fiuviatile, and the autumnal-winter occurrence of a large number of species during the height of the sewage contamination and bacterial development. The dominant species in this ciHate wave are Carckesium lachmanni, Epistylis, Amphileptus, Lionotus, Plagiopyla nasuta, Glaucoma scintillans, Stentor niger, and 5. cczruleus. Some species, as Halteria grandinella, have a wider seasonal distribution, and others, as Vorticella, Trichodina, Zoothafnnium, Pyxicola affinis, and many others, are adventitious in the plankton. Still others, as Rhabdo- styla, Cothiirniopsis vaga, Opercularia, and similar peritrichan parasites, are passive members of the plankton. The actively (9) 116 limnetic ciliates are very few. As such we may include Codonella cratera, Tintinnidium fiuviatile, and possibly Stentor niger. Car- chesium lachmanni and Epistylis enter the plankton only in the form of detached and often moribund zooids, and thus are not typical planktonts, though of quantitative importance in our plank- ton in the colder months. A large number of species not here reported occur in our collections made elsewhere than in the river channel, especially in places where the decay of large quan- tities of organic matter is in progress. This is not a condition normally found in the open water of lakes, though it may occur along their shores, where vegetation is found, or in regions of sewage contamination. In the waters of the IlHnois, on the other hand, the current, combined with sewage and industrial wastes and the organic detritus from the richest of fertile prairies, provides a suitable environment, even in the open water, for the support of a ciliate fauna of a magnitude somewhat unusual in fresh-water plankton. This fauna is present also in the back- waters, but is less abundant there than in the river itself. These species occur in greatest numbers of individuals in our plankton dur- ing the winter months at minimum temperatures, rising in November as the temperature falls below 50°, and declining again as it rises to this point in April. As shown by the bacteriological investigations of Jordan ('00) and Burrill ('02 and '04), the bacterial pulse attend- ing the decay of the sewage and wastes at Peoria does not reach Havana during the warmer months (see table on p. 231, Pt. I.), but when temperatures pass below 50° in November the increase in bacteria is marked. The decay is less rapid at low temperatures, and the process is still going on when the water in the channel passes Havana during the prevalence of low temperatures, and the ciliates that thrive in such an environment abound in the plankton at that time. The temperature limits of these ciliates of the period of bacterial development thus seem to lie between 50° and 32°. An examination of the plankton in the river at several points between Peoria and Havana at intervals throughout a year, will reveal how far the component species of this ciliate fauna are governed in their seasonal distribution in the plankton at Havana, respectively, by conditions of temperature and by the state of sewage contamination. The work of Roux ('01) upon the Ciliata about Geneva would seem to 117 indicate that many species of the fauna of stagnant water are more abundant in that region during the winter months. Owing to the difference in food conditions attendant upon the increase of sewage and bacteria during the colder months in the IlHnois River, it is impossible to determine from the data at hand the relative efficiency of the two elements of temperature and food in regulating the seasonal occurrences of our ciliates. Here, as elsewhere, the disastrous effect of sudden floods can be traced. The number of ciliates (Table I.) drops as floods rise, and recovers as the waters fall again. For this reason the winter occur- rences of the total ciliates are subject to considerable disturbances in the winter floods of the several years. The combination of the two methods of collection and of the two groups of ciliates, typical and adventitious, causes further irregularities (Table I.) in the sea- sonal distribution of totals. In the Illinois Ri\'er, for reasons given above, the Ciliata occupy a place in the economiy of the plankton of more than the usual im- portance. They feed principally upon bacteria, decaying organic matter, and the smaller algas, and are themselves eaten by the rotifers. I have found no evidence that they are utilized by the Entomostraca. They thus become active agents in the reduction of sewage and in the destruction of the bacteria of decay, in the purification of sewage-laden waters, and in the transfer of the matter in sewage to higher forms of animal life. The ciliates found in the Illinois include all the important species reported in the plankton of fresh water, and the list is somewhat larger than hitherto recorded in quantitative plankton collections in river or lake waters. These organisms escape readily through the silk net by reason of their small size, and in some instances the larger species, by reason of their mobility and flexibility, escape through the silk where less motile organisms of equal size are re- tained. By experiment I have found that well - shrunken silk bolting - cloth whose meshes average about 30-45/1 will not retain Paramecium whose diameter is 40-70/z. It may be that supple- mentary methods of collection which will correct the error of leakage will show that the Ciliata are of wider occurrence in the plankton than has hitherto been found to be the case. 118 DISCUSSION OF SPECIES OF CILIATA. Amphileptus spp.—Average number, 630. Amphileptus is a well- defined winter planktont in the river at Havana, and it affords a striking instance of the interdependency of organisms in the plank- ton. It feeds upon the heads of Carchesium lachmanni, engulfing the head in situ and encysting during digestion. Such heads, joined to the colony or free in the plankton, have been found in our waters. Its seasonal distribution at Havana is almost identical (Table I.) with that of Carchesium, upon which it feeds. Thus in 1897-98 Car- chesium was continuously present in the plankton from October 26 to Ma}^ 10, with a pulse, on December 7 of 283,800, and one on February 8 of 197,600. Amphileptus appears October 26 ; continues, with interruptions, to May 17; and has pulses December 7 and January 25, the latter reaching 13,545. In 1898-99 both appear early in October and have coincident pulses on November 22 and January 24. In 1895-9 6 the interdependence is even more striking, Carchesium reaching a greater development in this winter, with a pulse of 964,600 on November 27, and Amphileptus reaching 14,469 on this date and 14,835 a week later. Both species decline during the flood which follows, and rise during March to culminations, on the 24th, of 104,535 and 3,636, respectively. In 1898, Amphileptus disappears on April 12 at 52°, save for an isolated occurrence May 17 at 64°. It does not reappear until October 18 at 52°. In 1897, it reappeared October 26 at 59°, and in 1895-96 its limits were 45° and 48°, with the exception of one occurrence, April 17, at 66°. Carchesium occurs irregularly and sparingly during summer months, and Amphileptus was not taken in the plankton during that period. Its occurrence in the plank- ton is limited in the main to temperatures below 50°, but this limitation may be due primarily to the reduced numbers, at higher temperatures, of the organism upon which it feeds. It appears during the period of greatest sewage-contamination and bacterial development in the river at Havana. Roux ('01) ^nds Amphilep- tus most abundant in stagnant waters about Geneva in the winter months. Aspidisca costata (Duj.) Stein.—Found in the plankton but once —Jan. 11, 1898, at 32°. Bursaria truncatella O. F. Miill.—^Average number, 23. This large ciliate was found in the plankton at irregular intervals and in 119 small numbers. It was found six times in March ; twice in January and April; and once in February, July, and November. Its ap- pearance in the plankton is thus predominantly in winter months and at temperatures below 45°, though it occurs in the extremes of temperature conditions. Carchesium lachmanni S. Kent.—Average number, 26,546. This is normally an attached species, and its appearance in the plankton is due to the detachment of the heads. vSmall fragments of colonies are also found, but the greater number are isolated heads. The detachment seems to be a physiological process of the organism and not merel}^ the result of accidents. It is thus a detached and an adventitious planktont. Many of the heads taken in the plankton are in a moribund condition. For example, in a pulse of March, 1896, the following proportions were recorded. Date 120 October 1—that is, above 60°— are irregular and the numbers few (Table I.)- It is thus predominantly a cold-water planktont. Winter collections in 1894-95 and 1896-97 were too few to trace its seasonal movements. In 1896-97 it appeared November 5, rose to a maximum of 964,600 on November 27, and declined in the December-January flood (Pt. I., PI. IX.) almost to extinction, but recovered during its decline to a minor pulse of 16,160 on January 30. It again fell off in numbers during the floods of February (Pt. I., PI. X.), but rose during the decline of March to a maximum of 104,535 on March 17. Numbers become smaller and occurrences irregular after May 1. In 1897, Carchesium increased rapidly in late October to a small pulse of 13,200 on November 2, with a decline in the following fort- night, and a pulse culminating December 7 at 283,800, with subse- quent decline. The fluctuations during 1898 may be followed in Table I. The numbers increase during the slowly rising flood of January to a maximum of 197,600 on February 8 at 32°, and decline again during the more rapid rise (Pt. I., PL XII.) of the next three weeks. Stable conditions in early March bring about a pulse of 89,600 on March 15, and numbers decline again to 2,400 as the flood passes its maximum in the early part of April. As the levels fall another pulse of 99,200 appears April 26, from which a descent to minimum numbers—which prevail during the summer—takes place within a fortnight. The floods, especially sudden ones, seem thus to interfere with the appearance of Carchesium in the plankton, while gradual rises, as that of November, 1898, are not so detri- mental. The table of bacterial occurrences (Jordan, '00) in the Illinois at Havana and Pekin given on p. 231, Part I., indicates that the bac- terial development consequent upon the sewage and industrial wastes of Peoria extends down the river to Havana during the colder months of the year. The occurrence of Carchesium in the plankton is thus coincident with that of greatest sewage pollutionii and bacterial development at Havana. Carchesium is much more abundant in the channel of the river, where sewage pollution is greatest, than it is in the adjacent backwaters. It seems probable that the bacteria either directly or indirectly contribute towards its development, constituting, itmay be, an important element in its food. Flood waters, which dilute the sewage (cf. hydrograph and chlorine 121 in PL XLV. of Part T.) might for this reason tend to interfere with the development of Carchesium, and thus cut off the source from which the plankton individuals arise. I am not able, however, to trace any close correlation between the fluctuations of the chem- ical matters indicative of sewage and sewage decay and those of Carchesiiim. In the stable hydrographic conditions of 1897 we find a symmetrical pulse of considerable dimensions rising from 2,200 on November 9 to 283,800 on December 7, and declining to 26,500 on January 11, 1898. vStable low water with an ice blockade (Pt. I., PI. XI. and XII.) characterize this season. No explanation for the fluctuation is suggested in the physical environment. The chemical condition of the water, was, however, greatly disturbed (Pt. I., PI. XLIV.). The fivefold increase in free ammonia is indic- ative of approaching stagnation under the ice, and the threefold increase in chlorine marks the sewage concentration. Approaching stagnation might have caused the decline of Carchesinm, or it may be a specific reproductive cycle of the organism which combines with the external factors of the environment to produce such a wave of occurrence. Chilodon cucullulns Ehrbg.—Average number, 102. This species was found in the plankton in January and February during the bac- terial increase. It was also found in July. It escapes through the silk net, and does not ordinarily appear in plankton collections, though abundant wherever decay is active. Codonella cratera (Leidy).—Average number, 101,024 or 452,500*. This is the most abundant of the ciliates in our plankton, consti- tuting about one third of their total number. It appears in every month of the year, and in 1898 it was recorded in every collection but one, that of December 13 (Table I.). It is sub- ject to great fluctuations in numbers, its maximum occurrences tend- ing to appear in April, May, or June, and again in September or October. Minimum numbers prevail during the winter, when many of the shells are empty, and the midsummer interval is subject to pulses of varying amplitude. Spring pulses were detected as follows : in 1895, on April 29 (16,324) at 64°; in 1896, on April 24 (562,152) at 7 2°; in 189 7, on April 2 7 (470,000) at 60°; and in 1898, onMay 3 (73 6,000) at 60°. These vernal pulses coincide with or approximate closely to the dates of the spring volumetric pulses. This somewhat remark- able approximation of dates near the end of April may be the result, 122 in part at least, of the dates of collection ; but after allowance is made for this, the species still exhibits a seasonal cycle of remarkable regu- larity. The autumnal pulse is of less amplitude, and of less regu- larity in location as to time and temperature. In 1894 it appears September 4 (14,000) at 78°; in 1895, on September 12 (5,840) at 81°; in 1896, on August 29 (58,800) at 74° or October 14 (63,200) at 57°; in 1897, on October 5 (204,400) at 71°; and in 1898, on September 27 (92,800) at 73°. The midsummer pulses are, as a rule (Table I.), of less amplitude than the vernal or autumnal ones. In 1896 and 1898 exceptions to this statement appear in two large developments which follow in each case upon the decline of the June rise. In 1896 (Pt. I., PI. X.) this pulse (152,400) came June 11, and in 1898 (Pt. I., PI. XII.) it came (1,499,200) June 7 at 78° and exceeded in amplitude the re- corded vernal pulse. In both cases the pulse was recorded as occur- ring at an interval of a week after the crest of the June rise had passed. The character and sequence of these pulses is well shown in Table I. The occurrence of Codonella in abundance in the purer backwaters and in the plankton of our Great Lakes (Kofoid, '95) indicates that it is not dependent upon the sewage bacteria directly for food for its developmicnt in our waters. The appearance of the greatest pulses during a period of considerable sewage dilution still further indicates its independence of sewage bacteria. A comparison of the fluctua- tions of the totals of the chlorophyll-bearing organisms with those of Codonella affords some evidence of a correlation between the two. Of 39 pulses which can be traced in our records in the chlorophyll- bearing organisms, 21 precede and 13 coincide with those oi Codo- nella, while in the remaining 5 instances the multiplication of Codo- nella precedes that of the phytoplankton as a whole. Thus in the main the pulses of Codonella follow, or coincide with, those of the phytoplankton. The evidence of this sequence may be followed in Table I. by a comparison of the records of Codonella with those of the total phytoplankton. The sequence indicates that the food of Codo- nella may be found in the phytoplankton, and that these recurrent periods of growth have some connection with the conditions of nu- trition. The seasonal cycle of Codonella is closely followed by the other member of the family found in our plankton — Tintinnidium fiuviatile. 123 Codonella occurs throughout the whole range of temperatures. The Avinter minimum and the decline during the maximum temperatures of summer, combined with the presence of vernal and autumnal, or late summer, pulses, indicate that the optimum conditions for this organism lie neither in winter nor in summer. The spring pulse was at temperatures of 60°-72°, and the autumnal one at a wider range of 57°-78°. Permanent increase in numbers does not begin (Table I.) until March 1 5 at 46°, and the permanent falling off is found on November 15 at 41°. The optimum temperatures in our waters thus lie near 60°-70°, and conditions favoring growth are limited to a range of 10°-15° upon either side of the optimum. This species readily escapes through the silk net on account of its small size and its motihty, and such collections give at the best in- complete evidence of its seasonal distribution. The amplitude of its fluctuations is thus reduced, and owing to the irregularity of the error arising from leakage, the reduction is not proportionally distrib- uted throughout the year. Tests made of the loss of Codonella by leakage through the silk indicated that but one was retained to twenty-fourfound in the filtrate . Codonella was counted inboth the siIk and filter-paper collections, with the result that in 1897 the totals for the year (omitting one date on which the filter collection contained an unusually large number of Codonella) showed one Codonella in the silk to twenty-five in the filter collection. In 1898, however, the ratio was one to four and a half. The error in the filter collection is large, but data seem to justify the conclusion that only a small proportion of the Codonella is retained within the silk net. The proportion for the whole period of collection by the two methods (August 3, '97, to March 28, '99) is one to seven, if one date on which aberrantly large num.bers appear in the filter collections be omitted. This species is a typical planktont, and is apparently the same as C. lacnstris Entz, by which name it is designated by European writers. Leidy's name, however, has priority according to the accepted rules of nomenclature. It is an exceedingly variable organism, at least in the form, proportions, and size of the shell, in the degree of its con- striction, and in the foreign particles which fill its matrix. The rings or bands which ornament the orifice vary in their number, width, and relative proportions, and in the perfection of their developm.ent. The intergradation which these variants exhibit is sufficient to my mind to make their elevation to specific rank unjustifiable. 124 Codonella is an important element in the food of many of the Hm- netic rotifers, especially Asplanchna. Codonella is a common constituent in the plankton of our own Great Lakes (Smith, '94; Kofoid, '95; Jennings, '00a), and has been reported from most European waters. Apstein ('96) finds in German lakes major pulses in spring and autumn and minor ones in midsummer. Lauterborn ('94) reports Codonella in the plankton of the Rhine, and Schorler ('00) in that of the Elbe, but neither follows its seasonal history. Coleps hirius Ehrbg.—Average number, 13 . This species occurred in the plankton collections irregularly and in small numbers, princi- pally in autumn months during the height of the bacterial develop- ment. It escapes through the silk readily. Colpoda cucullus Ehrbg*.—Average number, 9,615. This species appears in the plankton principally during the colder months of bacterial predominance, from November to April, and occasionally during the summer. Cothurniopsis vaga (Schrk.) Blochmann was found in both 1898 and 1899 on Canthocamptus. Didiniimi nasutum (O. F. Miill.) Stein*.—-Average number, 12,692. This species also is found in the plankton during winter months, especially in November and December during the bacterial increase. It was also found in midsummer. Epistylis spp.—^Average number, 2,020. The free heads or frag- ments of colonies of one, or possibly of several, unidentified species of Epistylis, or it may be of Opercularia also, were associated with Car- chesium lachmanni in the plankton during the colder months, but in much smaller numbers (1 to 13 in 1898). Identification in most cases was impracticable, though in some instances E. flavicans Ehrbg. was determined, and it seems probable that most of the winter forms at least belong to this species. Hempel ('99) reports E. plicatilis on snails, and various other aquatic animals have been found infested with colonies of undetermined species of Epistylis. The distribution of Epistylis in the plankton (Table I.) is in its limits somewhat like that of Carchestum. It is more abundant and more continuously present during the period from November to June (at temperatures below 60°) than in the intervening warmer months. It is found throughout the whole range of temperatures. Its pulses coincide with those of Carchesium when they occur, but they are not 1 125 always found in Epistylis when they appear in Carchesium. This degree of similarity in the seasonal cycle of the two genera is indica- tive of their correlation with the same environmental factors, the principal one of which is the increase in bacteria attending the colder months. Euplotes charon (O. F. Miill.) Ehrbg. was taken but once in the plankton—August 23, 1898. Euplotes patella Ehrbg*.—Average number, 2,888. It was found in small numbers and at irregular intervals from April to December throughout the full range of temperatures. It was most frequently taken in the summer. Glaucoma scintillans Ehrbg.*—Average number, 39,615. This species was taken in the plankton from the middle of October till the middle of April. It was present in larger numbers and more contin- uously in December and February. It is thus a member of the plankton during the time of bacterial increase. Halteria grandinella O. F. Mull.*—Average number, 255,769. The seasonal distribution of this species in the plankton does not show the limitation to the winter months noted so frequently in other ciliates. It was found in every month of the year but May, in largest numbers in July and August, and most continuously in December and January. The data are too few and irregular to determine any pre- dominance as to season or temperature. Holophrya simplex Schew. was found in -small numbers in the filter collections of December, February, and March in the winter of 1896-97 at temperatures from 32° to 44°. Leucophrydium putrinum Roux.—Average number, 525. This species was recorded July-September, 1898, during the low-water period, at temperatures from 89° to 63°. It was described by Roux ('99) from stagnant water, but in our plankton no conditions of stag- nation attend its presence, though sewage contamination is great and decaying organic matter abundant. Lionotiis spp.—Average number, 94. With Amphileptiis in the winter plankton there occur a number of other, smaller, gymnostome ciliates which in best-preserved specimens resemble Lionotiis. A few occurring in March and April, 1898, were found to be L. fasciola Ehrbg., and it is probable that most of the individuals belong to this species, though exact identification is difficult with plankton mate- rial. The seasonal distribution of Lionotus coincides very closely 126 with that of Amphileptus. The species appear in November or De- cember and continue through March in temperatures below 50°, but the numbers retained by the silk net are too small to trace their sea- sonal routine. Their seasonal distribution in the plankton coincides with the period of greatest access of sewage and bacterial increase in the river at Havana. Roux ('01) finds this genus well represented in the fauna of swamps, and most abundant in October and March. Loxodes rostrum Ehrbg. was identified but once—March 22, 1897, at 44°. Nassula r^ihens Perty occurred July 30, 1897, at 84°. Opercularia articulata Goldf.—This species is parasitic upon aquatic Coleoptera. In the plankton of June 28, 1897, eleven colonies or fragments of a colony were found, the largest with 115 zooids. Opercularia nutans (Ehrbg.).—Average number of zooids, 60. In the plankton this species was found attached to Alona affinis in January, 1898, and to Cyclops in April and August. Opercularia not specifically determined were found free in the plankton in June and July ; in November, attached to Canthocamptus ; in January, attached to Brachionus—and even to the eggs of this species. An unidentified form was also found upon Cyclops. Ophryoglena atra Lieberk.—Five irregular occurrences of this species in small numbers were recorded in 1899 from January to the middle of March. Paramecium spp.—Average number, 41. Paramecium was found 18 times in the plankton. Two of these instances were in May and August at temperatures of 64° and 79°, and the remainder were between November 20 and March 30 at temperatures below 48°. Most of the occurrences are in midwinter at minimum temperatures under the ice. P. aurelia (O. F. Mlill.) has been found in the river waters (Hempel, '99), but not all taken in the plankton belong to this species. Specific determinations are not easily made with accuracy in preserved plankton material. In our plankton, Paramecium is present principally during the period of greatest contamination by sewage. Plagiopyla nasuta Stein*.—Average number, 1,181,000 during the winter of 1898-99 from November 29 to March 28. This species was not recognized in the plankton of previous winters. It reaches a pulse of 11,520,000 on January 3, 1899, at 32.2° under the ice. 127 Levander ('94) finds it in numbers under the ice in Finnish waters. On account of its motility and small size it readily escapes through the silk net. Pleuroncnia chrysalis (Ehrbg.) Stein.—Average number, 9. Re- corded only in January, 1898, at minimum temperatures. Prorodon farcius Clap, and Lach.—Only a few scattered occur- rences—from the last of September to the first of March at tempera- tures from 73° to minimum. An unidentified species of Prorodon was also found irregularly from November to April. Pyxicola affinis S. Kent.—^Average number, 58. This species is usually attached to aquatic plants, especially to Lemna. It has been found in the summer plankton from June to August during maximum temperatures, especially in 1896, when recurrent floods brought much Lemna from the backwaters into the ri\'er. It was found October 18 at 52°, attached to Melosira varians. Rhahdostyla spp.—Average number, 110. Peritrichan ciliates re- ferred to this genus have been noted on Cyclops, Canthocamptus, OligochcBta, and even in considerable numbers upon the body, append- ages, and eggs of Polyarthra platyptera. They have appeared thus passively in the plankton during winter months from December to March, especially in 1899. Stcntor ccnrideus Ehrbg.—Average number, 882. This species presents a characteristic seasonal distribution in our plankton. Its numbers are never very large, and its full cycle can not always be tracedt in the records. It is a planktont of the colder season in our waters. But three records—one July 28, 1896, at 82°, one August 3 of the same year at 80°, and a third, August 15, 1894, at 84°—lie outside of the period between September 1 and May 1. In 1898 (Table I.) the autumn "cycle begins September 6 at 79°, but in both 1895 and 1897 the species does not appear until late in November or in December at 34° or below. In years prior to 1898 the numbers were small and irregular, but on January 21, 1898, the maximum number of 28,800 was reached at 34°, under the ice, during the slowly rising flood of that month (Pt. I., PI. XII.). It accompanied an increase in Sientor niger, and there are indications elsewhere that the two species may fluctuate together. The high (Pt. I., PI. XLV.) chlorine (38.), nitrites (.175), and free ammonia (4.6) at the season of greatest development in the plankton are in- dicative of conditions approaching stagnation. The appearance of 128 this species in stagnant water has often been observed. Roux ('01) finds it especially abundant in September, October, and February in stagnant waters about Geneva. Stentor niger E^hrbg.—Average number, 3,124. In our waters this species also is a winter planktont (Table I.). There have been but four records of occurrence between May 1 and September 1 . In 1895-96 the species appeared November 14 at 44° and reached a maximum of 68,635 December 18, after three weeks of minimum temperatures and approaching stagnation under the ice. Numbers declined in the December-January flood (Pt. I., PI. X.), but rose again in March, as the flood declined, to 39,087 on the 24th at 40°. It disappeared from the plankton April 30 at 70° and did not re- appear until November 1 7 , from which time it continued until March 22. In 1897-98 it returned September 21 at 71°, attained a maxi- mum of 42,000 November 23 at 43°, declined during December, and rose to 47,000 on January 21 at 34° under the ice, and in the con- ditions approaching stagnation described in connection with the dis- cussion of S. ccbtmUus. a decline in num.bers continued until April 12 at 52°. Favorable conditions for growth are thus found in our waters between 32° and 50°, and the optimum seems to lie near 40° or below. This species reaches its greatest development in our waters during the time of greatest sewage pollution and bacterial development. It is known as a bog-water species, and was found by Roux ('01) in stagnant waters about Geneva during the colder months. Hempel ('99) reports this species as 5. igneus (?), but from the descriptions of Roux ('01) I am inclined to consider it as 5. niger Ehrbg. It may be that both species are included in our data, but they are predomi- nantly of the niger type. They include also individuals of the black- ish variety 5. igneus var. fuliginosus Forbes, which, it would seem from Roux's description of these species, should be transferred to 5. niger. The fuliginosus form was very abundant in the margins of Pine and Round lakes, Michigan (Kofoid, '95), during the summer in surface temperatures of 61°-70°, where sewage contamination was but slight. Stentor polymorphus (O. F. Miill.) Ehrbg. was found sparingly in July and August during maximum temperatures. Hempel ('99) reports S. barretti Barrett and 5. roeselii Ehrbg. from the river, but I have not identified them in the plankton collections. 129 Strombidium viride Stein was found in small numbers in January- March, 1899, at minimum temperatures. Stylonychia mytilus (O. F. Mtill.) Ehrbg. was found in the plankton sparingly from September to Februar}^, and once in June. Tintinnidiurn fliwiatile Stein.—Average number, 22,590 or 1,640,- 192*. This species is somewhat sharply limited to the warmer months in its seasonal distribution. In 1898 (Table I.) it makes its appearance April 4 at 49°, reaches a maximum of 720,000 May 3 at 60°, and has three decreasing pulses ; one of 104,000 on June 14 at 80°, one of 95,200 on August 2 at 79°, and one of 22,400 on September 27 at 73°, and disappears from the plankton October 18 at 52°. The records in previous years are more irregular, though traces of vernal and midsummer pulses can be found in the records. Filter-paper catches indicate that only one in eighty of this species is retained by the silk. They also locate the pulses as approximately coincident with those of the silk collections. Apstein ('96) finds Tintinnidiuni to be a spring planktont with its maximum in April in Lake Plon, while Seligo ('00) finds it in lakes near Danzig in the autumn, with a maximum in September. In our own waters in 1896 the autumnal pulse in August-September exceeds the vernal one. The gelatinous lorica of this species is subject to great variation in its size and proportions, and especially in the region about the aper- ture. A somewhat thimble-shaped form was described by Hempel ('96) as T.illinoisensis, the specific distinctions being based wholly on the lorica. This form intergrades with the typical lorica of T. fluviatile Stein, and should not in my opinion be given specific rank. Trachelius ovum Ehrbg.—Average number in 1895, 847. This' species did not occur in 1898 but was rather common in November- December, 1895, reaching a maximum of 10,695 on December 4 at 32.5°. Isolated appearances in small numbers in December and January of other years have been recorded. In our waters it is thus a winter planktont. Stagnation conditions under the ice were approaching (Pt. I., PL XLIII.) when the pulse of 1895 occurred in the Illinois River. Apstein ('96) found it, however, in Lake Plon with a maximum in May-June, disappearing in the summer and returning again in November. Trichodina pedicidiis Ehrbg.—Average number, 1; in 1897, 874. This species is normally found upon Hydra, on the gills and skin of 130 amphibians, and on young fish. It appears in the plankton during the summer months in every year except 1898, a single record only being made in that year. The earliest record was on June 1 1 , and the latest on November 3 1 . The whole temperature range is practically included in these occurrences, though the species disappears within a few weeks after the temperature falls below 50°. It usually appears in small numbers and irregularly, and no pulses like those of typical plank- tonts can be traced. A free life in the plankton is apparently not its usual habit. Zacharias ('00) has recently called attention to its appearance in the plankton in German waters. Vorticella rhabdostyloides Kell.—Average number, 61. This little Vorticella is found attached in small clusters to AnabcBna spiraides and occasionally to other members of the phytoplankton. It is some- what common in the waters of Lake Michigan, but is rare in spring months in the Illinois River. Vorticella spp.—Average number, 7,843. At irregular intervals from April to November isolated individuals and small clusters at- tached to bits of debris in the silt were taken in the plankton. They were most abundant at temperatures above 50°. The irregularity in their occurrences indicates that they are adventitious in the plank- ton. Identifications of plankton material are impracticable except in strongly marked species. Hempel ('99) has found V . campamila Ehrbg., V. microstoma Ehrbg., and V. similis Stokes in the river and its adjacent waters. Zoothamnium arhuscula Ehrbg.—A few colonies were taken in August and September in 1896 in the plankton, probably adventitious during the disturbed hydrograph of that year (Pt. I., PI. X.). The preceding list of 45 species does not complete the catalog of the ciliate constituents of the plankton, though it includes all of the species of quantitative importance during the years of our operations. The residium of unidentified ciliates, which, excluding the partial identifications in the above list, does not often exceed two per cent, of the total individual ciliates, includes principally isolated individ- uals of species difficult of identification or others whose preservation did not permit it, and a considerable number of small ciliates and of forms ectoparasitic upon Entomostraca and other planktonts. Most of these organisms are either adventitious or passive members of the plankton, and further study of the littoral region, of stagnating 131 waters, and of these parasitic forms will reveal the great richness of the ciliate fauna in this aquatic environment. SUCTORIA. Average number, 332. This class is not quantitatively im- portant in the plankton, being represented, in so far as our records go, only by adventitious or passive planktonts. No limnetic species has as yet been found in the Illinois. An examination of the littoral region during the prevalence of ciliates will probably yield a rich suctorian fauna. DISCUSSION OF SPECIES OF SUCTORIA. Acineta linguifera Clap, and Lach.—This species is usually found on aquatic Coleoptera. A single occurrence of an unattached indi- vidual was recorded June 21, 1898. Metacineta mystacina Ehrbg.—Average number, 301. This species occurred in the plankton from March till October in 1898 and in the winter mionths of 1899, at irregular intervals and in small numbers (Table I.). Most of its occurrences attend flood invasions, and it is evidently adventitious. It is frequently attached in the plankton to minute particles of debris. This species varies greatly in the size of the lorica. Sand ('01) gives the range in height as from 33-700 /<. The variation in proportions has given rise to a number of descriptions of new species by Stokes ('88 and '94) and Maskell ('87), but an examination of a series of individuals such as appear in the plankton shows that they intergrade so closely that specific distinctions can not be maintained for the variants. Meta- cineta appears throughout the whole range of temperatures, no seasonal predominance appearing in the records. Podophrya fixa O. F. Mull.—Average number, 12. This species is also adventitious in the plankton. It was recorded in March and September at 37° and 73°. Cysts were noted January 21. Tokophrya quadripartita Clap, and Lach.—Average number, 4. Adventitious in the plankton in March and November. Hempel ('99) finds it most abundant in May and June, associated with Epistylis plicatilis and Opercularia irritahilis on crayfish, insect larvae, and turtles. Tokophrya cyclopum Clap, and Lach.—Found occasionally upon Cyclops during spring and summer. (10) 132 PORIFERA. Spongilla spp.—Average number of spicules, 772. The identifi- cation of fresh-water sponges by isolated spicules is practically impossible, and, moreover, the sponge fauna of the Illinois River is as yet practically unknown. No attempt, therefore, was made to identify the species to which the spicules which occur in our plank- ton collections belong. They belong to the genus Spongilla in part, and were usually the simple sarcode forms, the gemmules or their spicules not appearing in the plankton. They occurred in all months of the year, and were found in 46 per cent, of the collections. They are adventitious, and their occurrence in the plankton is there- fore dependent in part upon hydrographic conditions. Records in December and January are few (3) and always occur on rising floods. In February and March, months of rising floods, they are increased (8 and 7), but decline again in April-June (3, 5, and 5), months of predominantly declining Vv^ater and more stable conditions. In midsummer and autumn mionths (July to November) they again occur more frequently (8 to 12), probably as a result of proximity to the season of greatest growth and frequency of sponges in the river and its backwaters. Here also they occur m.ost frequently in years of greatest hydrographic disturbance, as, for example, in 1898. The adventitious relation which they bear to the plankton is also seen in their erratic and irregular numbers. The maximum record (16,000 per m.^) was made June 28, 1897, on the rising flood; the next in size, on August 10 in stable low water. In both instances the plankton was probably taken from water in which as a result of some local disturbance the remains of some disintegrating sponge had been distributed. Living sponges are found in considerable abundance on submerged brush and timbers in the channel and backwaters during the summer months, and feed on the smaller organisms of the plankton, being one of its depleting agencies. CCELENTERATA. Hydra fiisca L.—Average number, 39. Hydra occurred in about 16 per cent, of our channel collections—a percentage which would be considerably increased if the whole of each collection had been examined for it, or if backwater collections should be included. With one exception the 28 occurrences recorded, all fall in May-September 133 at temperatures rarely below 70°. The earliest record in channel waters was on May 1, 1896, at 68.75°, and the latest on November 15, 1897, at 47°. Of the 28 records in channel waters the months from May to September have, respectively, 6, 3, 10, 7, and 1 record, and there is 1 in November. Hydra is thus a late vernal and a summer planktont in our waters. Observations in the field and a cursory examination of the col- lections made in the backwaters have indicated that Hydra is often very abundant on the vegetation. It is also limnetic in habit, floating with the foot attached to the surface film and tentacles widely extended; or, without attachment, in the deeper strata of water. A similar limnetic habit was often observed in the case of Hydra in channel waters, especially on still warm days when the surface was unruffled. Hydra was generally more abundant in the plankton in May or in early summer. The maximum record in channel waters was 3,200 per m.^ on July 21, 1897, the error of dilution being, however, large in this record. In Qwver Lake on May 8, 1896, a maximum record of 5,335 per m.^ was made, the error of dilution being very small. This was during a vernal plankton pulse (8.14 cm.^ per m.^) in these waters, when the food of Hydra was present in considerable abundance. Hydra viridis L. was seen frequently in spring-fed backwaters and in laboratory aquaria, but was never recognized in plankton collections made in channel or backwaters. The limnetic habit noted in H. fusca was not observed in the case of this species. PLATYHELMINTHES. TURBELLARIA. Numerically and from the volumetric standpoint the Turhellaria are not of great significance in the plankton of fresh waters as a rule. However, in some seasons and under certain conditions Stenostoma becomes very abundant, as, for example, in autumn months in back- waters, and generally where decaying vegetation abounds. In the autumn of 1895 the plankton in the relict pools of Flag Lake consisted almost entirely of Synura uvella, Stenostoma leucops, and Entomos- traca. 134 The average number in channel waters is 103 per m.^, and, as might be expected, their occurrences are erratic in seasonal distri- bution and their numbers are irregular. They occurred in channel waters in every month of the year and throughout the whole seasonal range in temperatures. The numbers in 1898 were larger and occur- rences more frequent in May, during the run-off of the spring flood, and smaller and more erratic during the rest of the year. In the total of all collections enumerated the percentage of occurrences was highest in June (60 per cent.), July (83 per cent.), August (48 per cent.), and October (47 per cent.), and lowest in colder months, when it rarely rises above 30 per cent. ' The numbers are also larger in the warmer months, a maximum record of 19,250 per m.^ on September 4, 1894, following a slight rise in river levels at low stages. The adventitious character of the Turbellaria in channel plankton is sug- gested by the erratic data, but the adaptability, at least of certain species, to the limnetic habit under certain conditions is also indi- cated by the large numbers. The identification of the Turbellaria in plankton collections is not feasible in the course of the usual methods of examination of pre- served plankton. Accordingly no effort was made to identify the individuals occurring in our catches. Many of them were evidently rhabdocoele turbellarians, and of these probably many were Stenos- toma leucops. The genus Vortex was also represented. Mesostomum ehrentergii O. Schmidt was taken in small numbers on August 26, 1895, along the shores of the river in vegetation. This identification is that of Dr. W. McM. Woodworth ('97). Stenostoma leucops O. Schmidt.—Average number, 2 1 . By far the greater proportion of the turbellarians in our collections probably belong to this species. The statements made regarding the group as a whole therefore probably apply to this species. TREMATODA. Many of our predaceous fishes and other aquatic vertebrates are infested to an extraordinary degree by flukes parasitic in the intestine or other viscera. This, in conjunction with the fact that the fish markets are located in house-boats along the stream and their refuse generally cast directly into the channel, is sufficient to account for the . few adventitious adult distomes which have been noted in our plank- 1 135 ton collections. They have occurred singly in February and July, but were not identified. The free-swimming larval stages or cercaria of unidentified trem- atodes were also found singly in August, September, and October. Aspidogaster conchicola v. Baer, which occurs abundantly in the mantle cavity and pericardium of many of the UnionidcB (see Kelly, '99), which form great beds on the river bottom, was taken in an immature condition in the plankton on June 27. Cotylaspis insignis Leidy, likewise a parasite of the Untonidcu, associated with Aspidogaster but confined principally to the mantle chamber, was taken in the plankton on February 4. CESTODA. Tetrarhynchus sp. was adventitious in the plankton on June 27, and doubtless of similar origin to the adult trematodes above noted. NEMERTINI. Fresh-water nemerteans were definitely identified as such in the plankton on only two occasions, July 23, 1894, and March 22, 1897. They were doubtless adventitious—from the shore or bottom, where they are most abundant. NEMATELMINTHES. NEMATODA. The free-living nematode worms are predominantly shore and bottom forms, living in the midst of the decaying organic matter of the bottom ooze. In a habitat such as ours, where the quantity of this decaying matter is very great, the nematodes are correspondingly abundant, and, owing to the unstable hydrographic conditions, they find many opportunities of joining the plankton temporarily. Ac- cordingly we find that nematodes are met most frequently and in largest numbers in rising flood waters, when the bottom deposits of tributaries and the main stream are carried in channel waters as silt. Thus, in the month of March nematodes occurred in 13 of the 15 collections examined, with an average number per m.^ of 465, while in August they were found in but 8 of 21 collections, and averaged only 186 per m.^ So, also, in the winter flood of 1895-96 nematodes were found in the plankton almost continuously till the middle of 136 April, while in the more stable conditions of the preceding year they were found in only one third of the collections. In 1897 most of the 31 collections examined were made in stable conditions, and nema- todes were found in but 5 of these, and 4 of these 5 were made in rising flood waters. In 1898, a year of greater hydrographic dis- turbance, nematodes occurred in 31 of the 52 collections, averaging 318 per m.^' to 82 in 1897. Of the 31 occurrences in 1898 all but 6 were in recent flood waters. The hydrographic conditions attending the presence of nematodes in the plankton thus indicate that they are adventitious in the plankton. Further evidence of this is to be found in their erratic numbers. Thus, on February 20, 1896, none was re- corded, and on the 25th their numbers rose in flood waters to the maximum record for all of our collections—18,422 per m.^ No effort was made to determine the species of these nematodes. A considerable variety of forms awaits the labors of some courageous systematist. ACANTHOCEPHALA. These worms are found abundantly in the Catostomidcs and other limophagous fishes of the Illinois River, and in many of the water- fowl which feed in its waters. A chance occurrence of a single specimen in the plankton on August 3, 1896, is probably to be ac- counted for as in the case of other intestinal parasites. ANNULATA. OLIGOCH.^TA. The representatives of this order belong to the smaller aquatic species—generally littoral or limicolous forms found especially in decaying vegetation or among LemnacecB, and belonging principally to the family Naididce—and usually occur in the plankton in mutilated condition, since autotomy occurs when the preservative is added to the plankton. Specific identification of the fragments is therefore often impossible and usually of questionable certainty. I am indebted to Professor Frank Smith for assistance in such identi- fications as have been made. The following list (see Smith, '00) gives the relative frequency of the species from which accessions to the plankton are made, with my notes on identified forms in the plankton. 137 NAIDID.-E. Stylaria lacustris (L.).—Abundant. Taken in the plankton in April. Nais elingiiis O. F. JMiill.—Abundant. This species was identified in the plankton on April 29, 1895, during the decline of the spring flood. Slavina appendiculata (D'Udekem) {Nats liirida Timm.).—Fre- quent. Ophidonais serpentina (O. F. Miill.) {Nais serpentina O. F. Miill.). —Frequent. Dero limosa Leidy.—-Abundant. Dero obtusa D'Udekem.—Abundant. This species was taken in the plankton in July and August, 1895, during the run-off of impounded waters from recently invaded back- waters. (See Pt. I., PI. IX.) Dero vaga (Leidy).—Abundant. Two individuals (Part I., p. 297) were found in channel waters in stagnation conditions under the ice on February 23, 1895, at a time when the plankton was almost entirely exterminated. Under normal conditions we have no evidence that this species is more abundant in stagnant waters. Dero fitrcata Oken.—Frequent. Pristina leidyi Smith.—Abundant. Pristina flagellum Leidy.—One specimen. ChcBtogaster limncei v. Baer.—Abundant. ChcBtogaster diaphanus Gruith.—Abundant. Chatogaster diastrophus Gruith.—This is apparently the most abundant species of the order in the plankton,—having been identi- fied in all months but January and May,—especially at times when impounded flood waters are drained off from backwaters, as, for example, in the March flood of 1895. ^OLOSOMATID^. Molosoma hemprichii Ehrbg.—Frequent. Molosoma tenebraruni Vejdovsky.—Abundant. jEolosoma sp.—Abundant. For reasons assigned above, the great majority of the oligochastes in the plankton remain unidentified and are included in our records 138 of total oligochastes. These records throw some light on the condi- tions controlling the occurrence of oligocheetes in the plankton and their seasonal distribution. They occur in all months of the year and throughout the whole seasonal range of temperatures. They appear in the plankton most frequently and in largest numbers in disturbed hydrographic condi- tions. Thus, of the 31 collections made in 1897, only 6 contained oligochastes, and the average number per m.^ was only 32. Five of the 6 collections containing oligoch^tes were made during the run-off of flood waters from impounding backwaters. In 1898, a year of much disturbed hydrograph (Part L, PI. XII.), there were 52 col- lections, in 35 of which oligochsetes occurred with an average number of 76 per m.^ Over 50 per cent, of the non-occurrences of oligo- chaetes fall in the more stable conditions of January, July-August, and December. The seasons of run-off from impounded backwaters are in all years favorable to the occurrence of oligochastes in the plankton. This is in sharp contrast with the nematodes, which appear with rising floods and access of tributary waters. The oligochastes are thus largely adventitious, at times when run-off from vegetation-rich backwaters prevails, and when Lemnacecs and Ceratophyllum are washed into the channel by hydrographic changes. ROTIFERA. (Plates III. and IV.) Average number, 592,416, of which 195,326, or 3Z per cent., are eggs, free or carried externally by the parent. Records were kept of males, of females, of females with eggs, of attached and free, summer, winter and male eggs, and of parasitized and dead indi- viduals. Rotifers occur in every collection and at all seasons of the year. Numbers are uniformly low (below 75,000 per m.^ and often below 15,000) during minimum temperatures from late in December till early in March. At other seasons of the year numbers fluctuate greatly, rarely reaching the level of the winter minimum except occasionally at the depressions between pulses. The curve of seasonal occurrence falls into the form of recurrent pulses (PI. III. and IV.) previously noted for other organisms. Of these pulses the vernal one in April-May is uniformly high, attaining 3,954,920 per I 139 m.3 on April 24, 1896, 2,287,160 on May 25, 1897, and the maximum record of all years, 5,247,800, on May 3, 1898. Pulses in excess of 1,000,000 per m.^ occur 14 times in our records: in July, August, November, and December in 1895; in April, 1896; in April, May, September, and October in 1897; and in May, June, August, Sep- tember, and October in 1898. There is, apparently, in years or seasons best represented in our records, a tendency for a vernal pulse, often the maximum one of the year, to occur in April-May, and for an autumnal pulse of large amplitude to appear between the last of August and the middle of October. The pulses contiguous to these major pulses of the year are often of considerable magnitude ; as, for example, in 1897, when the maximum of September 7 (5,121,- 000) is followed by another large pulse on October 12 (2,906,400), and in 1898, when the vernal pulse of May 3 (5,247,800) is followed by a June pulse, on the 21st, of large amplitude (2,601,200). The recurrent character of the pulses appears throughout maximum and minimum periods, and may be traced in Plates III. and IV. In the period of 15 months from July, 1895, to October, 1896, there are 10 such pulses, and 6 months in which pulses do not appear. In the 21 months from July, 1897, to March, 1899, there are 18 pulses, and 3 months in which they do not occur. They often coincide with or approximate those of the Entomostraca (PI. III. and IV.) and of the chlorophyll-bearing organisms (PL I. and II.). With the exceptions of the November-December pulses of 1895 at 33° (1,595,359 on November 27 and 1,636,640 on December 11) and the pulse of October 25 (1,048,620) at 48°, no pulse of con- siderable amplitude is found at temperatures much below 60° in channel waters. In the discussion which follows, 104 forms are listed, 6 belonging to the Rhizota, 6 to the Bdelloida, 91 to the Ploima, and 1 to the Scirtopoda. RHIZOTA, The Rhizota by virtue of their fixed habit are represented in the plankton either by adventitious species, torn from their location on water plants or other aquatic substrata by disturbances in the water, or by colonial species with a free-swimming habit, such as Conochilus. As represented by the latter type they are of some quantitative im- 140 portance in the plankton, especially of the backwaters. Average number, 8,796. DISCUSSION OF SPECIES OF RHIZOTA. Apsilus lentiformis Metsch.—An Apsilus doubtfully referred to this species was taken December 25, 1895, and April 29 and July 23, 1896, at temperatures of 41°-78°, in each case with rising river levels. Conochilus dossuarius Hud.—^Average number of females, 517. This species was more than ten times as abundant in the collections of 1896 and 1897 as in 1898. Hempel ('99) reports it from January to September, with a maximum in March. In the plankton collections of 1896 I did not record it until June 11, at 73°. It reached a maxi- mum of 25,800 July 18, and another of 142,800 August 15, at 86°, and disappeared from the plankton September 30 at 58°. In 1897 it reappeared May 25 at 66° and reached greatest numbers September 7 at 80°, and was not recorded after the 14th. In 1898 (Table I.) it first occurred March 8, at 37°, and attained its greatest number, 14,400, on September 27 at 73°. In 1899 it returned towards the end of January, under the ice, and continued till the cessation of operations in March. It thus occurs in the Illinois throughout practically the whole range of seasonal and thermal conditions, but not continu- ously. Colonies are of few individuals, and isolated individuals are often found in the preserved plankton. Females with 1-4 eggs were taken, and were most numerous during the rise of the pulse. About 4 per cent, of the females observed, were carrying eggs. Males were found on the decline of the pulse of July, 1896. Conochilus unicornis Rouss.—Average number of females 8,208; eggs, 100. Recorded from March 15, at 46°, to July 5, at 80°. A pulse of 8,000 on April 26 and one of 392,000 on June 7 constitute the only fluctuations. It was not found in 1897, and only sporad- ically, during the summer, in 1896. Females with 1-3 eggs attend the rise of both pulses in small numbers. The colonies of this species also are composed of but few individuals. Conochilus volvox Ehrbg.—Average number of females, 129. A few large colonies were taken March 29 and April 5 at 49°. Megalotrocha alboflavicans Ehrbg.—Colonies of this species are found in numbers on Ceratophyllum in the backwaters, and in 1894, 141 when the vegetation was common along the margins of the stream, it was taken in the plankton occasionally. Megalotrocha spinosa Thorpe.—Isolated individuals of this un- usual species were taken in small numbers in the plankton in August, 1S96, at maximum temperatures, but no colonies were observed. This is one of the largest of the rotifers in the plankton, individuals measuring about 1 mm. in length. The species was described by Thorpe ('93) from Chinese waters; was next reported by Weber ('98) from the neighborhood of Geneva, Switzerland ; and its occurrence in the Illinois is, I believe, the third record of its appearance. It affords another illustration of the cosmopolitan nature of the fresh- water plankton. In both Chinese and Swiss waters it was associated Avith M. semihullata Thorpe, also from Hong Kong and Brisbane. This latter species occurs in our waters also (Hempel, '99), though it was not taken in the plankton with M. spinosa. BDELLOIDA. Average number, 7,807. They were less numerous in 1897, a year of more stable hydrograph, and fully twice as abundant in 1896, when river levels were much disturbed during summer months. In their seasonal distribution, save for the increase of Rotifer tardus in the winter of 1898, the bdelloid rotifers reach their greater numbers in the plankton in the period from March to November. There is a trace of a vernal pulse in April-May (Table I.), and some irregular summer fluctuations, attributable in the main to floods. Their tem- perature optimum seems (except in the case of R. tardus above noted) to lie above 50°. They are as a rule adventitious in the plankton, owing their presence in some cases to floods, though the vernal in- crease can not in most cases be attributed directly to this disturbance. The species are difficult to identify in preserved plankton material, and the list here cataloged is small. Examination of living plankton would considerably extend the list of forms. DISCUSSION' OF SPECIES OF BDELLOIDA. Philodina citrina Ehrbg. was found in the plankton but once — Septem.ber 14, 1897. Philodina megalotrocha Ehrbg.—Average number of females, 351. This species was found in the plankton (Table I.) from March 15, at 46°, to November 8, at 45°. The distribution in previous years fell 142 within these limits excepting a single record December 29, 1896, at 35". The lower temperature limits are thus near 45°, and the num- bers are all small below 60°. The occurrences are never in very large numbers, and significant pulses do not appear—an indication that the species is adventitious in the plankton. The relative numbers in different years is suggestive. In 1896, with a total movement in river levels of 45.7 feet, the average number per collection is 770; in 1897, with a total movement of 44.8 feet, the number is 271 ; and in 1898, with 67.2 feet, it is 351. In 1896 a much greater proportion of the change in levels took place (Pt. I., PI. XI.) during the summer, when P. megalotrocha is present. With this in mind, it is apparent that a disturbed hydrograph tends to increase the number of this species in the plankton. A comparison of the individual occurrences (Table I.) w^ith contemporaneous conditions of the hydrograph (Pt. I., PI. XII.) in 1898, and in previous years also, shows that most of the larger records were made in planktons from a rising river. For example, the largest record made—8,000 on September 27, 1898 —is on the crest of a slight rise (Pt. I., PI. XII.). Some, however, appear in stable conditions, and may be attributed to the other causes of disturbance of the bottom and littoral fauna which tend to bring its constituents temporarily into the domain of the plankton. Rotifer neptunius Ehrbg.—Average number, 425. This species was found in the plankton in every month of the year but February, and thus throughout the whole temperature range. Between Novem- ber and March the records are scattered and the numbers small, while it is continuously present in larger numbers from March (50°) till late in October (50°-60°). The optimum temperatures thus seem to lie above 50° in our waters. The largest numbers recorded (22,224, April 29, 1896, at 72°, and 6,400, May 17, 1898, at 64°) attend the vernal volumetric pulse. Aside from this season, well-defined and symmetrical pulses are rarely traceable in the small numbers recorded. Some of the larger records, for example that of July 28, 1896 (10,200), attend rapidly rising water, but dependence generally upon this, agency for presence in the plankton is less directly evident in this species than in the preceding. As also in the case of R. tardus, the average number (246) in 1897, a year of more stable hydrograph (Pt. I., PI. XL), is greatly exceeded by that in 1896 (2,323), when the hydrographic conditions during summer were much disturbed(Pt. I., PI. X.). 143 Rotifer spp.—Average number, 199. Some bdelloid rotifers un- identified because of the state of their contraction, or not even questionably referable to other species listed, are here included. It is quite probable that some individuals belonging to the genera Philodina and Callidina are among the number. The occurrences are irregular. They exhibit a distribution with respect to years similar to that noted in the two species just discussed. Vernal pulses are noticeable in 1896 on April 29 (19,446), on April 27, 1897 (28,800), and ]\Iay 3, 1898 (3,200). Egg-bearing females were noted in the winter months of 1899, in December and March of the preceding win- ter, and in April, 1896. Individuals parasitized by Endophrys rotatoriorum Przesm. (?) were noted in April, 1896. , Rotifer tardus Ehrbg.—Average number, 6,688. This is the most abundant of all the bdelloid rotifers in our plankton, outnumbering all the others in 1898 six to one. This was due to a sporadic and un- usual pulse of individuals in the plankton in midwinter under the ice in 1898. Owing to this, the average number in 1898 exceeds that in previous years. If, however, the large numbers in January and Febru- ar}^ 1898, be reduced to normal winter proportions—no record in 1896 in this season exceeds 7,000—the average for the year falls to about 3,500. The average of occurrences in the plankton for 1896, 1897, and 1898 would then be 5,201, l,254,and 3,500, which approxi- mates somewhat the ratios of the relative disturbance of the hydro- graph in these years (Pt. I., PI. X.-XII.). The agency of flood water in affecting the numbers of this species in the plankton is to some extent indicated by this ratio. It is also apparent on compari- son of the seasonal distribution (Table I.) w^ith the hydrograph for 1898 (Pt. I., PI. XII.). The large numbers of January, February, and March appear in every case with rapidly rising water, and the same is true of the numbers on August 9 (12,000) and September 13 (17,- 500). Other disturbances than those due to floods, or other factors than disturbances in the water, must be invoked to explain such in- creases as one to 12,800 in April-May, 1898 (Table I.). This attends the vernal volumetric pulse (Pt. I., PI. XII.), but does not conform to its proportions. It appears in the more stable conditions of de- clining flood, and no adventitious factor is apparent to account for its development to such numbers in the plankton. The winter pulse was attended by large numbers of ovigerous females, but none was recorded during this vernal pulse. A somewhat similar increase 144 in stable conditions was found in March and April, 1896, from 40°-72°. Temperatures of 50°-70° were several weeks earlier than usual this year, but the increase in R. tardus came at lower temperatures than in 1898. As above stated, this winter pulse, or, rather, sequence of three pulses (Table I.),' culminating January 25 (89,397), February 15 (27,000), and March 15 (19,200), came with floods. No such increases attended the somewhat similar hydrographic conditions (Pt. I., PI. XIII.) of 1899 nor the winter flood of 1896. There is nothing in the eviron- mental data to explain this unusual occurrence. An unusually large number of females with eggs still attached to the body were seen in the period from January 21 to April 12. Fifty per cent, were ovigerous, carrying a single egg. Numbers of similar free eggs were also noted. Rapid multiplication of the species at the time of these pulses is thus suggested, and these may be dependent upon favorable conditions of nutrition of whose nature no clue is suggested. The species is in the main adventitious,with insufficient evidence of a par- tially limnetic habit at some seasons. The species occurs in almost every one of the plankton collections, and thus throughout the whole range of temperatures and environ- mental conditions. The largest numbers were taken during minimum temperatures under the ice ; but large numbers also appear at other seasons, and no temperature optimum is definitely indicated, though in years prior to 1898 the larger numbers and more regular occur- rences are to be found in the period from March to November at temperatures above 50°. Rotifer vulgaris Schrank.—-Average number, 275. This species has a seasonal distribution—though in smaller numbers and fewer occurrences—which corresponds somewhat closely with that of R. neptunius (Table I.). The same factors in the environment are pre- sumably operative in modifying its appearance in the plankton. Hempel ('99) finds R. macrurus Schrank, Philodina macrostyla Ehrbg., and Callidina elegans Ehrbg. in the plankton of Quiver Lake, adjacent to the river. PLOIMA. Average number, 571,611, including eggs, which constitute about 30 per cent. These rotifers occur at all seasons and are found in every collection. They are quantitatively the most important order 145 of the Rotifera. They include about 97 per cent, of the individuals and almost all of the limnetic species. As a group they exhibit a seasonal routine which is a complex of the records of individual species, and as such it reflects to a remark- able degree a similarity to individual records, especially of the peren- nial species. In general the Ploima are less abundant in colder months, that is, below 50°-60°, than in the warmer ones from May to October. Midwinter numbers are nevertheless considerable,—5,000 -35,000,—and with the first rise of temperature in March we have, in 1898, a pulse of 175,000 which declines and again rises in a vernal pulse of April-May, which vies with an autumnal pulse for rank as the annual maximum. Following the vernal pulse there comes a series of summer movements which vary from year to year. In 1898 they grow smaller as the season wanes, rising again in September. In 1 897 the autumnal pulse is the largest of the year and appears early in September. In 1895, on the other hand, it is carried into the last days of November. Numbers sink to the winter minimum shortly after the winter temperatures are reached. In a general way the direction of movement in the several parts of the seasonal curve of the total Ploima is much like that of the individual species of which it is composed. The differences lie in the amplitude of the pulses and in slight changes in the locations of maxima and minima. There are, it is true, many exceptions to this sweeping general statement, but it is, nevertheless, both surprising and significant that the sum of so many complex records should still preserve the recognizable outlines of its parts. This is not due simply to the dominance of a few abundant species, but is a combination of many, as will be seen frequently in Table I., where species with insignificant numbers still show in their seasonal occurrences some correlation with the movement of the great mass of the totals. This similarity points to some common factor in the environment common to all of the species. It is to be found, I believe, in the food relations—in the wax andwane of the food supply. Most, if not all, ploiman rotifers are herbivorous, or at least omnivorous, and find their food to a large extent in the phytoplankton. I have already called attention to the recurrent pulses of the chloro- phyll-bearing organisms. These primarily, but combined with other and largely changing seasonal factors such as hydrograph and tem- perature, are the basis upon which the superstructure of the seasonal changes in the ploiman plankton are built. The correlation between tn O O w o g o J o X o o 05 B to J d . 147 the seasonal distribution of individual species and these recurrent plant pulses will be discussed in connection with the various species wherever the data are available. For the present it will suffice to call attention to such correlation as exists between fluctuations of the phytoplankton and the total Ploima. The table on the preceding page gives the location and amplitude of the maxima of the ploiman pulses, and a graphic presentation of the seasonal curve of distribution of the total Rotifera will be found in Plates III. and IV. On comparison of the ploiman pulses with those of the chlorophyll-bearing organisms, graphically presented in Plates I. and II., it will be found that 15 of the 33 pulses of Ploima contained within the period covered by the plates coincide in location with the plant pulses ; that 12 follow at the next collection, usually a week later, and 3 within a fortnight ; while only 3 of the ZZ exhibit no such correlation. The data suggest strongly the agency of the plant pulses in building up the Ploima, and that the food relations are fundamental in the fluctuations of these planktonts. DISCUSSION OF SPECIES OF PLOIMA. Anurcsa aculeata Ehrbg.—Average number, 1.839. In 1898 this species has a very well-defined and characteristic seasonal distribution (Table I.). It first appears March 8 at 37°. increases to a maximum of 45,200 on May 10 at 61°. then declines, and disappears June 14 at 83°. The curve of its occurrence in this year is a very symmetrical one. It reappears on December 27 at 32°, and there are scattered occurrences through the winter months of 1899, Records in other years suggest in the main a similar distribution. In 1896 it first appeared January 6, rose to a pulse of 6,550 on May 8 at 76°, and, on the decline of the June rise, there was a second and larger pulse of 29,600 on June 17 at 76°. It reappeared on Decem- ber 29, and in 1897 reached a vernal maximum of 22,400 on May 25 at 66°, then disappeared, and was not again noted in the following winter nor until March 8. In 1894 the last vernal record was made June 12, and on September 4, at 78°, there was an autumnal pulse of 13,825—a phenomenon not repeated in subsequent years. The normal course of its seasonal distribution in the river plankton seems to be as follows: reappearance in December when minimum temperatures have been reached; slow multiplication during the winter, and a well-defined pulse on the decline of the spring flood in (11) 148 April-May with the possibility of a second on the June rise; and prompt and complete disappearance when maximum summer tem- peratures are established. Low water in the autumn seems to inter- fere with an autumnal pulse. In 1894 there was a well-sustained rise in September (Pt. I., PL VIII.) and a pulse of A. aculeata. In 1896, however, no pulse occurred in the high water of the autumn. No midwinter occurrences followed the very low water of 1897. It is thus in channel waters a vernal planktont, with its temperature optimum near 70° but below the summer maximum." Hempel's statement ('99) that it is a "winter species" is borne out by its presence from December through the winter, but its numerical distribution ranks it at once with the vernal organisms. Lauterborn ('94) finds it abundantly in winter months in the Rhine, and Ap- stein ('96) speaks of it as a " Sommerform," absent from Lake Plon from November till March, and with maxima from April to July in different bodies of water where it continues through the summer and till October, and then disappears. Summer temperatures in these Waters, however, are not recorded by him above 21° C. (69.8° F.), which is about the temperature at the time of the vernal maxi- mum in the Illinois, and at least 10° F. below that of the summer maximum in our waters. Jennings ('94, '96, and '00) records it as abundant in the summer plankton of Lake Erie, Lake Michigan, and some inland lakes of Michigan. These waters also are somewhat cooler (5°-10° F.) than those of the Illinois River in midsummer. Temperature, it seems, must have a decided effect upon the seasonal distribution of this organism in our waters, though the chemical conditions and food supply may also enter as factors in the summier suppression of the species. Females carrying usually a single egg appeared in 1898 early in April, and were most abundant during the maximum of the pulse. On an average, less than a fourth of the females were ovigerous. Empty loricae appeared May 10 (4,800) and 17 (3,200) at the crest and decline of the spring pulse, and the same phenomenon of deca- dence was noted in previous years during this period. Outbreaks of parasites were not recorded for the species, and the decline is to be attributed to cessation of reproduction and to the death and destruction of the individuals by the more usual causes. This species is quite variable, but no effort was made to follow its seasonal history. The type form is by far the most abundant. 149 A. aciileata var. valga Ehrbg. was seen frequently. A. serrulata Ehrbg., regarded by Weber ('98) as a variety of A. actileata, was recorded Jan. 24, 1899, and found by Hempel ('99) in December. It seems to be rare in our plankton. Forms approaching A. aculeata var. hrevispina Gosse were also noted, but they, too, are rare, being recorded only in February and March, 1899. A. aculeata var. curvicornis Ehrbg. was noted April 29, 1896, at 70°. AnurcEa cochlearis Gosse.—Average number, 69,393, distributed as follows: A. cochlearis {sensu strictu) together with A. cochlearis var. macracantha Lauterborn, 9,421; .4. cochlearis var. tecta Gosse, 15,432; and forms with posterior spine of intermediate length between cochlearis and tecta which include A. cochlearis var. stipitata Ehrbg.. 44,540. Numerically this is one of our imiportant species, containing over one ninth of all the rotifers in 1898. It is surpassed only by Brachionus bakeri (with varieties included), Polyarthra, and SynchcBta. Average number of eggs, 32,358. This is a perennial planktont, appearing in every month of the year throughout the whole range of temperature. Its entire absence in August, 1898 (Table I.), is not paralleled in any other year. In 1897, for example, there is a well-developed pulse of 45,600 on August 24. In 1894, 1895, and 1896 there is a midsummer minimum of a few weeks' duration in July, August, or September, but it is irregular in its location. While the appearance of sexual cycles was not traced by the records of males and winter eggs,—a matter of some difficulty and uncertainty in preserved plankton material,—the existence of such cycles is suggested by the recurrent pulses of occurrence in this species (Table I.). It is possible that the species is polycyclic in our waters. The pulses in 1898 are well defined, in fact, somewhat better than in previous years. The following table gives the num- bers in the pulses in the several years and the dates and tempera- tures at which the maxima occurred. All of the large pulses save those of November and December and one at the close of October (Oct. 25, 1898, 28,500) lie at tempera- tures above 60°. The vernal pulse of April-May is the largest and appears between 60° and 70°, and the amplitude diminishes as the period of maximum heat progresses, though in 1898 there was a recurrence of larger numbers as temperatures fell. The optimum 150 Pulses of Anuria cochlearis. Year 151 the distribution of the pulses with reference to the floods and the appearance of pulses during rising water suggest the operation of other factors than the one arising from contribution from back- waters. The pulse must be dependent to a large extent upon food supply of the organism, and a correlation between its periods of multipHca- tion and the pulses of its food, the chlorophyll-bearing organisms, is to be expected. A comparison of the seasonal distribution in 1898 (Table I.) and the pulses of chlorophyll-bearing organisms (PL II.) reveals the fact that three of the A. cochlearis pulses coincide with those of the plants constituting their food, and the other three coincide in part only, the remainder of the chlorophyll-bearing groups reaching their culmination a week prior to that of the rotifer. In 1897 the three pulses of A. cochlearis which lie in the common period (PI. II.) all culminate a week (in one case in part in fourteen days) after the maximum of the plants in question. In 1896, three pulses coincide and three follow in the subsequent collection; and in 1895, two coincide and two follow. Collections at daily intervals would be necessary to follow the correlation more accurately. It is probable from these juxtapositions and sequences in the A. cochlearis-algce pulses that we are dealing with a food relation. Multiplication of algse leads to increase of AnnrcEa, which, in turn, reduces the alg«,and then itself declines until the food planktonts again increase. Anurcca cochlearis is exceedingly variable in the length of the posterior spine, in the development and degree of curvature of the anterior spines, in the arrangement of the areas of the lorica, and in the degree of its ornamentation by small spinules. The separation of these varieties where every individual must be assigned to some one of them, is a matter of some difficulty owing to the presence of intergrading individuals. The characters which signalize var. kispida Lauterborn and var. irregularis Lauterbom are not quickly rc>cognized under the conditions of rapid plankton enumeration, and no effort was made to trace their seasonal distribution in our plank- ton. Lauterborn's var. macracantha was included with the type form—his var. typica—in our records. These two include those individuals with medium-sized and longer posterior spines. In our waters the variety macracantha is relatively rare, at least as figured by Lauterborn ('98). Indeed, both the type and this variety consti- 152 tute less than a seventh of the total representatives of the species. Their distribution throughout the year (Table I.) accords with the results obtained by Lauterborn ('98), who found that the average length of the posterior spine from January to May and from October to December was from 78 to 48 /it, while from June to September it was from 28.5 to 21//. In Table I. it will be seen that the longer- spined forms which I have referred to A. cochlearis var. macracantha and var. typica occur in the plankton from January to May 3 1 , and then disappear, returning again, in small numbers, October 25. The short-spined variety referred by me to A. cochlearis var. stipitata and the spineless var. tecta are, on the other hand, continued during the summer. The natural result would be that the average length of the spines in the species as a whole would fall during the summer months. It is apparent that this tendency on the part of A. cochlearis to become shorter and smaller during the summer months does not bear out the contention of Wesenberg-Lund ('98) that winter individuals are smaller and summer ones larger among perennial rotifers. He reports var. tecta as "die Hauptform des Winters " in several Danish lakes, and the variety with a long horn as a summer form, found in July-August. Of these varieties, macracantha, typica, and stipitata intergrade in our waters with numerous connecting links, while var. tecta is not connected with the other forms by many individuals with inter- mediate characters. Lauterborn ('98) also notes the greater inde- pendence of this variety in the waters of the Rhine. In Table I. the 'seasonal distribution of these three varieties, the long-spined {typica and macracantha), the short-spined {stipitata), and the spineless {tecta) are given separately. It will be noted that the long-spined form has the distribution above mentioned, that var. tecta runs throughout the whole year, and that var. stipitata is absent in midwinter and is a common summer form. The relative numbers of the varieties fluctuate in different years. For example, var. tecta was relatively but one fourth as abundant in 1897 as in 1898. As shown in Table I., whenever coincidently present in the plankton all the varieties respond to the causes which produce the rhythm of occurrence, the rise, culmination, and decline of the pulses being much alike in all of the varieties. About three eighths of the females noted in 1898 were ovigerous, carrying as a rule but a single egg. Instances of two eggs were 153 noted, but they are rare. The greatest proportion of egg-bearing females appears during the rise of the pulse, as is seen in the follow- ing table, which gives the data of the vernal pulse in 1898. From Anur.^a cochlearis. Date April 12. April 19. April 26. No. of Dvigerous females 800 6,400 45,000 May 3 536,000 May 10 i 489,600 May 17 110,400 May 24 6,000 May 31 3,000 Total females Total Ratio of eggs to individuals 2,200 i 800 15,200 8,800 I 137,800 65,000 1,022,400 552,200 1,145,600 643,200 \ 434,800 ' 160,000 i I 21,200 7,200 11,200 I 3,400 2.75 1.73 2.12 1.85 1.78 2.71 2.94 3.29 No. of dead 400 3,200 9,600 99,200 100,000 1,800 1,800 April 12 to the crest of the pulse on May 10 (not inclusive) the aver- age ratio of eggs to individuals was 1 to 1.87. From the crest to the foot of the decline inclusive the ratio is 1 to 2.98. The number of empty loricLe is given below', and it wdll be noted that on the week prior to the crest of the pulse there were 107 living to one dead; on the crest itself, one to twelve ; while the week following the crest of the pulse there w^as an empty lorica for every 4.3 living females. Rapid multiplication thus attends the rise of the pulse and rapid destruction its decline. Parasites w^ere very rarely observed in this species. The decline of a pulse is thus due to the cessation of reproduction and a relatively heavy death rate. Apstein ('96) finds that in Lake Plon Amtrcsa reaches its maxi- mum in July and is at its minimum in April. It is ever}rwhere common in the German waters. A. tecta, on the other hand, was found only in the smaller lakes and in great numbers, replacing cochlearis in w^armer months to some extent. Lauterborn ('98) regards it as the most abundant rotifer in the Rhine. Our statistical records do not show that this is the case in the Ilhnois, for it is here 154 surpassed by several other species. Zimmer ('99) finds that this species is the most common winter rotifer in the plankton of the Oder, with a maximum in the spring and a predominance of var. tecta from July to September. Schorler ('00) finds it to be the most common rotifer in the Elbe—from April to November ; and Skor- ikow ('97) finds it in the TJdy, in Russia, throughout the summer in great numbers, but surpassed by SynchcEta, Polyarthra, .and Brachionus angularis. The variety tecta greatly exceeds var. stipitata in these waters. Seligo ('00) finds it throughout the year in Prussian lakes near Danzig, with a maximum in May. There are indications, in his data, of recurrent pulses during the summer, but his interval of collection is too great to follow their history. Burck- hardt ('00a) finds it throughout the year in vSwiss waters, with its single maximum in August. Jennings ('94, '96, and '00) reports it in the summer plankton of Lake Michigan and Lake Erie and of inland waters of Michigan. AnurcBa hypelasma Gosse.—Average number of females, 2,390; of eggs, 1,917. This species has a very definite limitation to a period extending from early in June to the first days of November. There are but two records outside of these limits—a single female and egg on Jan. 11, 1898, and another upon April 19 of the same year. The probabilities of occurrence in very small numbers at all tempera- tures is thus indicated. The following table gives the data of pulses and temperatures. All of the pulses save one occur at temperatures above 70°, and with this exception the species declines rapidly and disappears shortly after temperatures pass below 60°. It is plainly, in our waters, a summer planktont, with its optimum temperature close to the summer maximum. This species takes no share in the vernal pulse, and there is no satisfactory evidence of any fluctuation corresponding to it at any other season. There are three or four pulses in each summer, and the species is apparently polycyclic, for winter eggs were found in 1898 either at the maximum of the pulse or the week or fortnight following. Thus 24,000 winter eggs were recorded on Sept. 27, 1898, the date of the maximum of the Septem- ber pulse. The parthenogenetic eggs preponderate during the rise of the pulses in a very marked manner in this species. For example, in this September pulse 55,400 eggs were recorded during its rise to 500 during its decline. In like manner, in the case of the 155 Pulses of Anur.-ea hypelasma. Year 156 but none appears in this interval, while hypelasma runs a normal course of recurrent pulses throughout the summer. This August pulse of hypelasma (Table I.) culminates August 16, just a week after the symmetrical and well-defined pulse of chlorophyll-bearing organisms (PI. II.) of August 9. With a single exception, all of the pulses of 1896 and 1897, indi- cated in the table, fall a week later than, or coincide with, the pulses of chlorophyll-bearing organisms, as in 1898. This species has not occupied a prominent place in the literature of fresh-water plankton. Weber ('98) finds it rare in Swiss waters in the summer. Lauterborn ('93) classes it with the monocyclic summer forms in the plankton of the Rhine, though he states in a footnote that he had found winter eggs once in June. It is probably polycyclic in our waters. Skorikow ('96) finds it in the summer plankton of the river Udy, in Russia, but it is not mentioned by other investigators of the potamoplankton of Europe. Apstein ('96) does not report it from Lake Plon. Asplanchna hrightwellii Gosse.—Average number, of adults 2,079, of eggs, 396; averages in 1897, 16,161 and 2,156. This is a poly- cyclic perennial planktont in our waters. It has been found in every month of the year, but the greater numbers and more con- tinuous occurrences lie between May 1 and October 30. In 1898 (Table I.) all but 200 of the 108,120 recorded, lie within these limits, and all but 260 above 60°. In previous years approximately the same limits are found. The following table gives the data of pulses and temperatures. Pulses of Asplanchna brightwellii. Year 157 Pulses of Asplanxhxa. brightwellii—Continued. Year 158 ASPLANCHNA BRIGHTWELLII. Date Males Females without eggs Ovigerous females Winter eggs May 3 " 10 " 17 " 24 " 31 June 7 " 14 " 21 " 28 July 5 " 12 " 19 " 26 August 2 . . . 9... 16... 23... 30... September 6 . 13 20 27 October 4. . . 11 .. 8,000 1,600 240 4,000 3,200 4,800 5,600 400 200 200 800 100 120 40 12,400 7,200 80 3,200 1,600 540 3,200 500 1,000 12,800 8,000 4,000 300 40 240 5,260 12,000 800 800 800 600 3 , 200 1 , 600' 100 200 400 120 60 5,600 800 800 60 1,600 1,000 500 159 An examination of the location of the pulses of Asplanchna brightwellii shows (Table I.) that in 1898 one coincided with the pulse of chlorophyll-bearing organisms (PI. II.) and the remaining four followed it either in a week or fortnight. In previous years two pulses coincide with and five follow those of chlorophyll-bearing organisms, and a single ill-defined one (Nov. 14, 1895) precedes. This species is not wholly herbivorous in its feeding habits. Codonella, Difflugia, and even other rotifers such as Brachionus and AnnrcEa,a.Te frequently seen in the digestive tract. Diatoms, even Melosira and PeridiniidcB, as well as Pediastrum and other algas, are frequently taken as food. In one instance a Daphnia cucullata 300 // in length was seen in the stomach in a transverse position. It was fully a third the length of the animal which had eaten it. Asplanchna brightwellii is reported by Skorikow ('97) in the summer plankton of the Udy, in Russia ; by Schorler ('00) as spo- radic in the Elbe in June and September; and by Lauterborn ('93) from the Rhine, where its cycle coincides with that of A. priodonta. Zacharias ('98) reports it in German reservoirs in June and August. It is a cosmopolitan species, but does not seem to have been found by other plankton investigators in European waters. Asplanchna ebbesbornii Huds.—Average number of adults in 1895, 942. In 1898, only winter eggs of the species were noted in the plankton in February, June, July, September, and October, though adults were doubtless there. Adults have doubtless oc- curred sporadically in all other years, and in 1895 reach a pulse of 21,518 on July 6 at 81°, which was followed by the appearance of males and winter eggs. All records of adults lie V.)etween April 29 and September 14 and above 60°. This rare rotifer has not appeared in the literature of fresh-water plankton elsewhere to my knowledge. Hempel's statement ('99) that his record of its occurrence in the Illinois is the first for this continent must be modified, since Leidy ('87) found it near Philadelphia. It is evidently a summer plank- tont in our waters, and the wide distribution of its winter eggs suggests that it, too, may be polycyclic ; and their appearance in the plankton in large numbers with reference to the adults taken, leads to the further inference that its center of distribution is prob- ably not in channel waters, and that it may be predominantly limicolous species, or have its center of distribution in the quieter backwaters. 160 Asplanchna ^irodi de Guerne is reported by Hempel ('99) in the backwaters in April. Asplanchna herricki de Guerne.—Average number, 15; in 1897, 295 ; in 1 896, 317. This species was always rather rare in our waters, and is apparently a summer planktont. The earliest record is April 29, at 64°, and the latest, November 15, at 48°. There is an indication of a vernal pulse in April-May in 1896 and 1897, and the recurrence of the species at intervals of a few weeks during the sum- mer suggests a polycyclic habit similar to that of other members of the genus in our waters, but the data are insufficient to follow the cycles if such exist. Ovigerous females were present when numbers were greatest, and males and females with winter eggs were found at the time of the vernal pulse on May 25 (3,200) in 1897. Hempel's statement ('99) of its rarity in June and July is not borne out by the statistical records in these months in 1896, 1897, and 1898. This rotifer is abundant in the summer plankton of Lake St. Clair, Lake Michigan, and lakes of northern Michigan (Jennings, '94 and '96), and it may be significant that it reaches its greatest devel- opment in the Illinois in the spring at 60°-70° and not during the period of maximum heat. This is about the summer temperature of those northern waters. This species has not to my knowledge appeared in the literature of European plankton, though it is found in European waters. Asplanchna priodonta Gosse.—Average number, 441 ; winter eggs, 7. This species is much less abundant in our waters than its associate A. brightwellii, being outnumbered by it five to one in 1898. It is in the Illinois River a summer planktont only, at least so far as the records go, though reported elsewhere as perennial. The earliest record in any year is April 29, 1896, at 70°, and the latest October 5, 1897, at 70°, when an unusual pulse of 22,000 was found. The records are too scattered to trace the seasonal history. There are only indications of recurrent pulses. In May, 1898 (Table I.), the best-defined pulse is recorded. The details, which conform in the main to the sequence noted in A . brightwellii of ovigerous females with summer eggs during the rise, with males and winter eggs at and after the culmination of the pulse, are given in the appended table. This is the only cycle found in this year. The presence of ovigerous females and winter eggs at other seasons as well, in other 161 ASPLANCHNA PRIODONTA. Date 162 is probably a polycyclic planktont with its greater pulses in spring and fall. Asplanchnopus myrmcleo Ehrbg.—Taken in small numbers and irregularly from May to October at temperatures above 60°. Ascomorpha ecaudis Perty.—Found rarely in early summer, in temperatures above 60°. Brachiomis. The discussion of the species of this genus in our plankton is fraught with great difficulty. The genus is represented in the Illinois River by a very large number of individuals (fully 25 per cent, of the total Ploima), and the species are, almost without exception, exceedingly variable. They are loricate forms, and the variations affect the proportions of the lorica and the development of its prolongations in spines, antlers, and various diversifications of its surface. They are evident upon the most cursory examination in most cases, and have been utilized by systematists for the estab- lishment of species. . For example, Weber ('98) lists no less than 67 species of Brachionus, the most of which he regards as synonyms, and he includes only a part of the species. Fuller knowledge of the extreme variability in this genus has led the most thorough students of the rotifers to regard many of these so-called species as but varieties at the best, and to express their opinion with unmistakable plainness that descriptions of new species among rotifers should only be made after most careful determination of the variability of the organism (cf. Rousselet, '02, Jennings, '00, Wesenberg-Lund, '00, and Weber '98). For one not a specialist in rotifers, the attacking of the Brachionus problem from the statistical standpoint is made difficult by the condition of the literature of the subject, owing largely to the semi- tropical distribution of the genus; by the absence of any critical monograph of the Vv^hole genus dealing fully with the synonymy of the subject; and by the necessity of establishing and maintaining constantly amid the ceaseless change of varying forms the same standards of distinction between the species or varieties into which all of the individuals enumerated must be assorted. Furthermore, these distinctions must be established before the plankton is counted ; that is, before the limits of variation are fully appreciated. It is needless to say that my efforts are at best but approximations 163 to a satisfactory analysis of the genus in our waters. Brachionus contains by virtue of variation in the hard parts of its lorica most excellent material for the study of the problem of variation, and its rapid multiplication makes possible a correlation with seasonal and environmental changes not often afforded. Evidence has accumulated in the various papers of Schmarda, Ehrenberg, Barrois, v. Daday, Anderson, and others who have dealt with the microscopical fauna in tropical regions, that this genus attains its greatest development in the warmer waters. It is there- fore not strange that Skorikow ('96) finds the genus well represented in the warm and shallow waters of Russia, and that the plankton of the Illinois River and its backwaters should contain a large and varied representation of the genus. For convenience in treatment I have arranged the individuals of Brachionus under the following species, without, however, intending to indicate thereby that they have equal claims for specific recogni- tion. The most of these include one or more varieties, and in desig- nating the varieties I have taken those forms—for exam^ple, in Brachionus hakeri—whose descriptions most closely fit the predomi- nant varieties in our waters, designating them often without com- plete consideration of all synonymic possibilities. In some cases several possible varieties have been included under one head. The following is the list of species with the varieties which have been thus separately enumerated. Brachionus angularis Gosse, " var. bidens Plate hakeri Ehrbg. " var. bidentatus Anderson brevispinus Ehrbg. cluniorbicularis Skorikow nielhemi Barrois and v. Daday ohesus " rhenanus Lauterborn tuberculus Turner budapestinensis v. Daday militaris Ehrbg. mollis Hempel pala Ehrbg. (12) 164 Brachionus pala var. amphiceros Ehrbg. " " " dorcas Gosse " " " " forma spinasus Wierz. " quadratus Rousselet " urceolarts Ehrbg. " " var. rubens. Ehrbg. " " " hursarins Barrois and v. Daday " variabilis Hempel Brachionus angularis Gosse.—Average number of females, 57,890; of males, 25; of summer eggs carried, 29,560; of winter eggs, 1,223 ; of male eggs, 54. Of the individuals, 13,973 belong to var. bidens and 43,942 to the type; of the eggs, 2,035 belong to the variety and 28,802 to the type. The combined statistics of the species will be' discussed before the type and variety receive separate treatment. This species was found in every month of the year and throughout the whole range of temperatures, but the period of continuous presence and large numbers lies definitely between May 1 and November 1 and above 60°. In fact, in 1898, 98.6 per cent, of all the individuals were found between May 3 1 and October 4 and above 70°. Approximately the same conditions are found in previous years save in 1896, when an earlier spring (cf. PI. X. and XII., Pt. I.) is attended by an earlier appearance of this species. Temperature seems thus to have a very decided effect upon the seasonal distribution of the species, and may have something to do with its apparent absence in the cooler waters of our Great Lakes and of L. St. Clair, for in spite of all the work done upon rotifers in those regions by Jennings it has been found but once—by KelHcott ('97) in a cove at Sandusky. This identification may be questionable, since he says "I at first took it for B. mollis Hempel." Notops pelagicus, since described by Jennings ('00), is found in the plankton of Lake Erie, and according to him this species is much like B. mollis in its appearance. In any event B. angularis is very abundant in our warm waters and practically absent in the more northerly waters of Michigan, whose summer temperatures are 10°-15° below that of the Illinois River and its backwaters. Brachionus angularis presents the usual phenomenon of recurrent pulses, but in spite of the large numbers they are rather less regular 165 than usual—for example, than those of Anu7wa (Table I.). This irregularity is somewhat more pronounced in the separated records of the type and variety (Table I.) than in their combined statistics. This fact that their combined curve of occurrence is more regular than their separated curves constitutes, to my mind, evidence that we are dealing only with one genetic cycle, and that the variety does not belong to a fully separated genetic series. The following table gives the data of pulses and temperatures in the several years. Pulses of Brachionus angularis including var. bidens. Year Date 166 near the summer maximum. The location of the pulses with respect to those of the chlorophyll-bearing organisms (PI. II.) shows in the main the same relation that has been observed in other ploiman rotifers. In 1895, three angularis pulses lie in the period common to both, one of these coinciding in location and two following at the next collection. In 1896, two coincide and five follow at the next collection or shortly thereafter. In 1897, four follow at an interval of a week or a fortnight, and one is located where data are incom- plete. In 1898, three coincide and three follow at a short interval, and one (June 7), a minor and ill-defined pulse, appears to lie on the rise of the pulse of the chlorophyll-bearing organisms. In the main the dependence of these rotifer pulses upon the recurrent periods of increase in these primal links in the food cycle is suggested by this coincidence or sequence. The pulses of Brachionus angularis co- incide in the main with those of the totals of ploiman rotifers (Table I.). There is no vernal pulse in the species at the time of the April- May volumetric maximum, and no large autumnal pulse. The pulses in August-September, at the close of our period of maximum heat, average much greater than those of other months, and still further indicate the relation of this species to the higher temperatures. The eggs are carried by the female attached to the posterior end of the lorica. Usually but a single summer egg is carried at one time, but often two, three, and even four, have been seen during the height of the period of rapid reproduction. The relation of the number of eggs to the pulses is obscured in this species to some extent by the fact that the eggs are similar to those of other Brachio- nus and when detached cannot be identified with certainty. Records are therefore based upon attached eggs only. The num.ber of these depends to some extent on the detachment in the processes of collection, killing, and subsequent handling. In a few cases de- tached male or winter eggs could be identified with some degree of probability by the constitution of the rotiferan plankton. An examination of the records of eggs (Table I.) will, however, suffice to indicate the prevalence of rapid reproduction during the rise of the pulses and the decline in the process during the fall of the pulse. Males, male eggs, and winter eggs were recorded in a number of instances at the culmination or during the decline of a pulse. For example, in 1898, they followed the pulses of August 16, September 167 6, and especially that of September 27, when they were found con- tinuously for a month. « The separate records of the type and the variety (Table I.) contain in their seasonal distribution one point of special interest ; namely, the appearance of the variety after the type has been present for some time. iVn examination of the records in the several years reveals the fact that var. bidens is practically confined so far as large numbers are concerned to the months of July-September. This appears in 1898 (Table I.) and is equally evident in 1896 and 1897, but is less noticeable in 1895. The first large pulse is passed in each year before var. bidens takes any appreciable part in the genesis of the pulses. Even the second large pulse is not extensively con- tributed to by the variety in some instances. On the other hand, the later pulses in 1895 and 1897 were mainly of the variety. There is thus in this species some evidence of a tendency on the part of the variety marked by the development of a pair of posterior spines to appear in the latter part of the period of seasonal occurrence. The variety bidens in our records includes individuals with well- developed spines {B. caiidatus Barrois and v. Daday), btit they are not to my mind worthy even of varietal distinction, since they intergrade so completely wdth var. bidens and are merely well-de- veloped examples of this variety, and I see no reason for giving the variety two names. Wesenberg-Lund ('00) has expressed the opinion that the elonga- tion of structural processes which he has noted in summer planktonts is an adaptation on their part to the changes in the buoyancy of the water dependent upon changes in its specific gravity and, as shown byOstwald ('03 and '03a), in its molecular friction caused by seasonal fluctuations in temperature. It would seem that this tendency on the part of the spinous form of Brachionus angularis to appear in greater proportions in late summer at the period of maximum heat in our waters might be an illustration of Lund's thesis and Ostwald's theoretical considerations. The changes in temperature during the occurrence of the species are, how^ever, not very great, though our incomplete records suggest (Pt. I., Table III. and PI. X.—XII.) that August temperatures are higher on an average than those of July. The averages for June, July, and August are 77.75°, 81.03°, and 81.49°. In 1897, the dominance of the spinous type extends well into September, but it accompanies a period of summer heat (Pt. I., 168 PI. XI.) prolonged for a fortnight into September, with river water at or above 80°. In 1898, it falls away in numbers more rapidly than the spineless form (Table I.) as temperatures fall in October, though this tendency is less marked in previous years. Brachionus angularis, as above stated, seems to be rare in the plankton of our more northerly and cooler American waters. It is also conspicuousl}'' absent from plankton of Swiss waters, as reported by Weber ('98) and Burckhardt ('00 and '00a), and from German lakes examined by Apstein ('96), Zacharias ('98), and Seligo ('00), and from Finland waters examined by Stenroos ('98). It was, however, found by Wesenberg-Lund ('98) in Danish waters, and in the Udy River, in Russia, by Skorikow ('97), whose statistical records show it to be the most abundant Brachionus in that stream, and outnumbered among the rotifers only by SynchcBta stylata and Polyarthra. Schorler ('00) finds it in the Elbe from April to July and most abundantly in June. Lauterborn ('98) reports it as perennial in the Rhine and polycyclic, with winter eggs in April, June, August, October, and November. This distribution is much like that in the Illinois River, and will probably be found in tem- perate waters wherever the seasonal cycle is thoroughly examined. Brachionus baken Ehrbg.—Average numiber of females, including all varieties, 594 ; eggs, 420. The following table, giving the average of each of the varieties in the several years, will serve to indicate their relative abundance, the totals showing the relative abundance of the bakeri group in each year and of each variety in the total of all the collections. Though the species is greatly diversified by variation the number of individuals is much less than that of many other plankton rotifers in which variation is much less apparent. It will be noted that the species was apparently more abundant in the earlier years. This is only in part the result of the distribu- tion of the collections, as is shown by the fact that the numbers taken were much larger. Thus in 1898 the largest record is 7,600 ; in 1897 there are three occurrences in excess of this; in 1896, two; in 1895, three; and in 1894, four. The largest occurrence, 122,958, was on June 30, 1894. The largest numbers by far were recorded in 1894, a year of low water in spring. The hydrographic conditions of the following year were somewhat similar, but the development of B. bakeri was much reduced, at least at the time of the collections. 63 < a p Z O X o < pq 170 The reducing effect of the recurrent floods of 1896 may be traced in the smaller numbers recorded in this year ; and the larger num- bers of 1897 may be referred to the more stable conditions then prevailing. The very small numbers of 1898 may also be due to disturbed hydrographic conditions of that year. The number is much smaller than in 1896, when the hydrograph was even more disturbed, but in this latter year there was more run-oft' of impounded backwaters during the occurrence of B. baken,SLnd this would tend to favor their appearance in channel waters. The occurrences and numbers of this species (as a whole) are ever3rwhere somewhat irregular, so that pulses of occurrence are somewhat ill defined. Several such pulses are indicated in 1898, and others recur in the records of previous years. As suggested by the data of 1898 (Table I.), the several varieties share in these pulses. The evidence upon this point is much more striking in other years, when numbers are larger. For example, in the following table note the pulse of 26,800 on August 23, 1897. Date 171 B. pala, B. angiilaris, and probably B. iirceolaris, has a variety—in fact, several varieties—with two posterior spines which are usually symmetrically placed but not always symmetrically developed. The form without posterior spines (var. dumorhicularis Skorikow) inter- grades with these, and a series might be formed with complete intergradations linking this in turn with var. rhenanus Lauterborn, in which the spines are but slightly and often unequally developed. From this we pass, by a slight elongation of the posterior spines, to var. hrevispinus Ehrbg., thence to the type in which the spines as figured by Rousselet ('97) are directed posteriorly with but slight curvature. From this we may pass toward variants in which the symmetry is preserved, but the spines are much elongated and curved outwardly. The anterior spines in such individuals are also more elongated and exhibit a similar outward curvature (var. melhemi Barrois and v. Daday). Extreme types of this curvature sometimes occur {B. falcatus Zach.). In another direction we find the bilateral s^^mmetry of the processes, both anterior and posterior, to some extent lost as a result of difiierences in the curvature of the spines (var. tiiberculiis Turner). There are also differences in the surface markings of the lorica which have been utilized as specific distinctions. Kertesz ('94) describes as B. granulatus a species with a minutely pustulate surface, and Turner's B. Uiherculus takes its name from this same feature. It seems questionable, however, if these surface markings are even of varietal value. Individuals without spines, in which the transverse diameter is relatively large (var. obesus Barrois and v. Daday) , are also found. In assorting the individuals belonging to this variable group I have arranged them imder the following heads: bakeri O. F. Miill., bidentata Anderson (non bidentatus Kertesz), brevispinus Ehrbg., duniorbiailaris Skor., melhemi Barrois and v. Daday, obesus Barrois and V. Daday, rhenanus Lauterborn, and tuberculus Turner. The number might have been increased. The individuals referred to \'ar. melhemi include many if not all of the long-spined specimens such as Rousselet ('97) has referred to the type, the latter designa- tion having been given to individuals intermediate between this and brevispinus. The variety tuberculus includes the asymmetrical individuals, regardless of the surface markings. I will now briefly compare the seasonal distribution of these varieties and note any peculiarities which mark them individually : — 172 Brachionus hakeri O. F. Mull., type form.—Average number, 2. As shown in table on p. 1 93 (MS.), this form is much more abundant in previous years though it is relatively rare, ranking sixth in the list of seven forms recognized. The most of the records fall prior to the middle of August, and it seems to be an early rather than a late summer form. Brachionus hakeri var. obesus Barrois and v. Daday.—Average number of females, 41 ; of eggs, 62. The proportion of egg-bearing to non-egg-bearing females—2 to 3 in all records—is larger than in any other variety. It seems probable that the lateral expansion which marks this variety may be only the result of rapid reproduc- tion. In common with most of the other varieties this one occurs at the time of the pulses, but it is last in the list of seven, and the numbers are too small to trace its seasonal preferences with cer- tainty. Brachionus hakeri var. hidentatus Anderson (non Kertesz). — Found once—August 5, 1895, at 78°. Brachionus hakert var. cluniorhicularis Skor.—Average number of females, 90; of eggs, 95. This also was more abundant in all previous years. This variety is, next to tuherculus, the most abundant of the varieties in our plankton. The two stand at opposite extremes of the series of varieties, the former being least modified, and the latter most, especially in the direction of asym- metry. It includes about one third of all the individuals of the species. The ratio in the grand total of females to eggs carried — 11,708 to 5,976—is somewhat less than the average in the entire species. This variety is distributed throughout the whole seasonal range of the species with no marked predominance in any particular part of it. It is wholly absent in the early summer of 1 897 , but very abundant in late summer of that year, though not in other years. The autumn of 1897 was one of long-continued high temperatures (Pt. I., PI. XL), and under those conditions this variety constituted two thirds of the individuals belonging to the species. If we add to it the representatives of rhenanus, obesus, and brevisptnus we have a total of 15,400 individuals with no posterior spines, or with spines but slightly developed, in contrast with only 2,200 with such well- developed spines referred to varieties melhemi and tuherculus. The conditions of temperature were those in which according to the 173 hypothesis of Wesenberg-Lund ('00) we should expect a predomi- nance of the long-spined forms. Brachiomis bakeri var. rhenaniis Latit.—Average number of females, 118; of eggs, 138; but more abundant in previous years. This is the third in numbers on the list of seven varieties, being surpassed only by cluniorhicularis and tuherculus. It includes about one sixth of the individuals referred to this species. It is found throughout the whole range of the seasonal distribution of the species and exhibits the same peculiarities noted in cluniorbicularis, to which it is very closely related. The proportion of females to eggs noted in this variety is very large; 5,284 to 5^485 in the grand total. Brachionus bakeri var. brevispinus Ehrbg.—Average number of females, 795 ; of eggs, 390; but somewhat more abundant in previ- ous years. It was found throughout the whole seasonal range of the species, but not quite so abundantly in the latter as in the earlier half of the summer, resembling in this particular the type. The nttmber of eggs carried in this species is in relation to the number of females less than usual—3,906 to 795. Brachionus bakeri var. melhemi Barrois and v. Daday.—Average number of females, 49; of eggs, 49. More abundant in previous years, especially in 1894, when it constituted over a fifth of the individuals (25,764) in the largest pulse recorded for the species as a whole—122,958 on July 30. In the aggregate in all years it includes only about a ninth of the individuals referred to the species. • This form was originally described from Syria, but it is found in great perfection in our plankton, even in the extreme type described by Zacharias ('98b) as B. falcatus. It occurs throughout the whole seasonal range of the species, its distribution being somewhat similar to that of tubercuius. I do not find any constant tendency limiting its occurrence to any part of the seasonal range. Brachionus bakeri var. tubercuius Turner.—Average number of females, 155 ; of eggs, 42 ; but very much more abundant in previous years, especially in 1894, when it constituted almost half (55,332) of the largest pulse of the species (122,958). This, the most divergent of all the varieties, constitutes over a third of all the individuals referred to the species. It occurs throughout the whole seasonal range of the species, though the larger numbers were found in 1894-97 in the earlier part or middle of the summer. I find nothing 174 in a comparison of the seasonal distribution of these more decidedly spinous varieties of B. bakeri with that of the smoother forms, such as cluniorbicularis, which indicates any correlation with temperature conditions of a nature to support Wesenberg-Lund's suggestion that the elongation of the processes of plankton organisms arises in response to the lessened buoyancy of the water during higher tem- peratures. Forms with and without such processes are found among the varieties of this species, and both occur indiscriminately through- out the whole range of seasonal occurrence, and, so far as I can see, the statistical data of their distribution with respect to temperature afford no evidence of a correlation of spinosity and high tempera- tures in this species. Other factors doubtless enter into this problem and obscure this response if it exists. B. bakeri is ever3rwhere widely distributed in fresh water. Its occurrence in the plankton of open waters has not, however, been a matter of frequent note. In fact there is some reason to think that it is largely confined to shallow warm waters where vegetation is close at hand, or where at least the flagellates and smaller algas abound, as they do in water fertilized by decaying vegetation or other organic matter. There is, it seems, no reason for regarding this species as merely adventitious in our plankton. It bears all the characteristics of a true limnetic organism in our environment. Its presence in the plankton is not due to floods or other disturbances which might carry it from a littoral region into the open water. It exhibits characteristic pulses, and is found everywhere in summer in company with typical planktonts in open water. Zacharias ('98) records it in some German ponds and streams, and Weber ('98) in Swiss marshes in the warmer months. Stenroos ('98) also finds it in the summer plankton of littoral and open waters in the shallow Nurmijarvi Lake in Finland. Jennings ('00) reports it as one of the commonest rotifers in East Harbor, Lake Erie, and in the swamps on the islands. In land-locked pools short-spined varieties were found, and in swamps the long-spined. Speaking of this difference, Jennings says "Possibly the different form found in these pools is due to the greater concentration of various salts in this water or to some kindred factor." In our own region both varieties occur at the same time in the same environments, channel and backwaters alike, and such factors as Jennings suggests to explain the appearance of the varieties cannot well be operative here in 175 channel waters. Schorler ('00) reports the species as sporadic in the Elbe, and Skorikow ('97) finds both B. hakeri and its variety hrevispinus sparingly in the Udy in summer months. This species in common with other BrachionidcE was infested by Bimosriiim hyalinum Przesm., and occasionally by a filamentous fungus-like growth. Empty loricae were wont to appear with the culmination of a pulse and subsequently. No males were identified as belonging to this species, and attached male eggs were recorded only late in September, 1897, at the close of an unusual pulse. They were found on var. cluniorhicularis and rhenanus. Females with winter eggs w^ere not at any time recorded for this species. It may be that some of the free winter eggs referred to the genus Brachionus (Table I.) belong to this species. The recurrent pul§es are similar to those of known polycyclic species, and we may infer the probability of such a phenomenon in B. hakeri, though conclusive proof of its occurrence is not found in the statistical records. Brachionus hudapestincnsis v. Daday.—Average number of females, 4,211; of eggs (carried), 740. This is one of the most sharply defined species of Brachionus and a typical planktont of open waters. It has, moreover, a sharply limited seasonal distribu- tion in which it is apparently polycyclic. The appended table gives the dates and temperatures of appearance and disappearance and the pulses in the several years. In the main, the period of occurrence is practically from the end of June till the early part of October and above 60°. A record in May, 1896, and an isolated one in December of the same year, indicate an extension of this period, but such occurrences are rare and irregular and the numbers small. This abrupt decline in 1898 as temperatures pass 60° (PL XII., Ft. I., and Table I.) is paralleled in previous years. The normal seasonal routine seems to be as follows: The species reappears in the plankton in May-June at 70°, rising slowly to its first pulse (average, 26,104) in July, with a larger pulse (average, 184,453) in the following month during the maximum heat, and a much smaller one (average, 10,044) in Sep- tember, followed immediately by an abrupt decline. The average temperature of the larger pulses lies close to the season's maximum, while the latest pulse at the lower temperature (72.2°) averages but 10,044. These data all indicate that this is a midsummer planktont, with its optimum temperature near the summer's maximum. The g en W Q Z < < , averaging 45. The temperatures on June 1 (Pt. I., PI. IX.-XII.) average about 75°, and on Oct. 1 about 67°. The spinous form (amphiceros) thus in- cludes about 45 per cent, of the individuals at low temperatures, and 93.5 per cent, at high temperatures; and the smoother form (pala type), 55 per cent, and 6.5 per cent., respectively. This predominance of the spinous variety at high temperatures is apparently a striking illustration from statistical evidence of the hypothesis of Wesenberg-Lund ('00) that such elongations of the body of planktonts are adaptations to the lessened buoyancy of the warmer water. This relation of the spinous form to higher tem- peratures is evident in every year, 1895-1898, and the proportion of spinous forms, 86-96 per cent., exhibits all the constancy that might at the best be expected in plankton data. The relation is generally apparent (Table I.) in the individual entries as well as in the sums total, and, considering the numbers concerned and the long period of observation, should have more weight than some of the exceptions to the hypothesis, which have been or w411 be noted, in which the data are less extensive. For example, Brachionus pala var. dorcas does not in its seasonal distribution support the hypothesis, but owing to its small numbers—especially of the form spinosus—less weight should attach to its evidence. In 1897 the first autumnal pulse of the pala group consisted almost entirely of var. amphiceros. This pulse started August 10 at 3,600, culminated August 31 at 1,398,000, and declined to 800 September 29. Of the 3,500,200 individuals included in this pulse, all but 1 1 ,400 belonged to amphiceros. The temperatures recorded during this period ranged from 83° to 71°. A second pulse started October 5 at 1,600, culminated October 12 at 1,605,600, and declined to on October 26. Of the total individuals (1,609,000) included in this pulse, 894,800 belonged to amphiceros and 7 14,200 to pala. The range in recorded temperatures in this period was from 71° to 59.5°. This may serve as an additional illustration of the relation of tem- perature to the spinous variety of Brachionus pala. This variety is itself polycyclic, as is evidenced by the recurrence of male and winter eggs carried by the female at times of the pulses. 191 Owing to the ease with which such eggs are detached, the records are quite imperfect indices of the actual numbers. In 1898 male eggs (carried) to the number of 70,400 per m.^ attended the culmina- tion of the vernal pulse (419,200) on May 3. Winter eggs (carried) were recorded twice on the decline of the pulse of August 1 6 ; once on the decline of that of October 25 ; and once on that of December 1 5 . Brachionus pala var. dorcas Gosse.—The seasonal distribution of this variety is so sharply defined that it merits especial attention. The following table gives the dates and temperatures of last and first records in each year. Seasonal Limits of Brachionus pala var. dorcas. Year 192 The curvature of the median anterior horns which defines this variety resuhs in a considerable elongation of these processes. With regard to the idea of Wesenberg-Lund ('00) that this tendency on the part of plankton organisms to elongate in " Balanceapparat " is an adaptation to the lessened buoyancy of the warmer water of summer, it must be said that it seems difficult to apply this hypothesis in the case of B. pala var. dorcas, which is probably a seasonal variety confined to winter months. I have no data, however, on the relative development of these processes in B. pala ^a.t different temperatures, beyond the seasonal limitation of this variety to lower temperatures when it should be least expected according to the hypothesis. Brachionus pala var. dorcas has not been found widely distributed in the fresh-water plankton, or at least not reported separately from B. pala, which is widely distributed. Skorikow ('96) reports it from Charkow, Russia; and Kertesz ('94), in January from Budapest. Brachionus pala var. dorcas forma spinosus Wierz.—Average number of females, 33 ; of eggs, 2. This form was always sporadic in its appearance in our plankton. Of 12 occurrences, 3 were in April, 2 each in November, December, and July, and one each in January, May, and August. The whole seasonal range of tempera- tures is thus included. It may be of significance for Wesenberg- Lund 's hypothesis that the spinous form of dorcas makes over 50 per cent, of its appearances between April 1 and- September 30, whereas dorcas itself is much less abundant relatively within these limits. The largest occurrence of spinosus—100,044 on April 24, 1896—was marked by the fact that 97.5 per cent, of the individuals were infested with fungi. The nearest approaches to pulses in this form are the November-December appearances in 1895 and 1896. Females with winter eggs were recorded December 29 in the latter year. Brachionus quadratus Rousselet.—Individuals corresponding to Rousselet 's description have been found occasionally in the plankton from the last of May till the middle of August at temperatures of 70° and above. The species is somewhat closely related to the bakeri series, and may ultimately prove to belong to it. Rousselet ('97) is of the opinion that it is distinct by reason of the truncate posterior end, the absence of foot sheath, the reticulations of the shell, and the semi-jointed foot. It occurred only in small numbers. 193 and forms intermediate between it and bakeri were not recorded. This is, I believe, the first record of its occurrence in American waters. Brachioniis urceolaris Ehrbg.—Average numl3er of individuals including all varieties, 468; of eggs, 56. The species was relatively quite abundant in 1897 (5,290 and 1,976) in the stable conditions then prevailing, but less so in the recurrent floods of 1896 (1,020 and 494). It is not a common species, being outranked by B. angularis, bakeri, budapestinensis, and pala. The species as a whole is found throughout the entire year, though never in large numbers since 1895. The following table, which gives the principal pulses in the several years, shows the wide range of the species and its varieties in seasonal distribution. Pulses of Brachionus urceolaris. Year 194 There is some tendency, especially in later years, toward the colder months. Eight of the fifteen pulses occur below 70°, and twelve between September 1 and May 1 . On account of the small numbers the pulses are poorly defined in our records (Table I.), but there are indications that they coincide in location, in a general way, with those of other BrachionidcB and the Ploima as a whole. They also in many instances coincide with or follow shortly after the pulses of chlorophyll-bearing organisms, as has been noted in other BrachionidcE. This species, B. urceolaris, is a cosmopolite, and of general occur- rence in the fresh-water plankton of smaller and warmer bodies of water. It is reported by Weber ('98) from Swiss marshes, by Zacharias ('98) and Marsson ('00) from many smaller German waters, and by Seligo ('00), throughout the year, from lakes near Danzig, where it attains maxima in April, July, and September. Since this author includes B. angularis (B. urceolaris forma angu- latus Seligo) with his records of urceolaris, it is probable that the species in the usual sense may have much more restricted numbers and range in his region. Kertesz ('94) finds it about Budapest. It is reported as sporadic in the vernal plankton of the Elbe by Schorler ('00), and is listed from the Oder by Zimmer ('99). Skorikow ('97) reports it once in summer plankton of the Udy near Charkow. The species is exceedingly variable in the development of the anterior spines, and in the proportions of the bod3^ It varies toward the bakeri group, and individuals are sometimes found'which seem to connect the two groups. I follow Skorikow ( '96) in placing B. rubens as a variety of B. urceolaris, including in it those forms whose anterior spines are least developed. The m.ore slender summer forms I have listed as var. bursarius Barrois and v. Daday. From my observations on B. variabilis Hempel, I am inclined to regard it as a possible variety in the urceolaris group. In form, texture, proportions, and anterior spines it is certainly similar to this group. The presence of the posterior spines would not suffice to separate it, since these may or may not be present, and the existence of a variety of urceolaris with such spines w^ould only present a phenomenon parallel to that observed in pala, angularis, and bakeri. The quadrate foot-plate present in variabilis, which, according to Hempel ('96), is not found in other species of the 195 genus, serves to distinguish this form, and in the absence of proof of its occurrence in forms of urceolaris as here defined I prefer to leave variabilis as a separate species. In any event it is closely related to the urceolaris group, and may ultimately be found to belong within its seasonal range of variation. Seligo ('00) has suggested that B. angnlaris is also a variety of urceolaris, but I do not so regard it. The averages of the different forms in the several years are given in the table on the next page, which also includes B. variabilis. The discussion of the different varieties follows : — Brachionus urceolaris Ehrbg., type.—Average number of in- dividuals, 18. The type form was not abundant in an}-' year, and its appearances were sporadic. It was recorded in February, June, and July. It includes less than one per cent, of the individuals referred to this species. Brachionus urceolaris var. rubens Ehrbg.—Average number of individuals, 244; of eggs, 41. This variety was more abundant during the stable conditions of 1897 (5,290 and 1,976) and the low- water years of 1894 and 1895. It includes over 99 per cent, of all the individuals referred to this species. It is apparently the winter form of the species. This appears clearly in its seasonal distribution in the later years, but in 1894 and 1895 it was found in summer months and in large num- bers. It is thus capable of development in the whole range of temperatures. The pulses recorded in the table on page 193 are in the main composed of this variety. It is quite abundant during the summer of 1894, attaining a pulse of 181,764 on August 15 at 84°, disappear- ing in September, and not reappearing until the April collection. It attains a pulse of 324,254 on June 19 at 80°, declines in July, then occurs sporadically until the following February. It then continues till June 6, with a pulse of 8,398 on April 17 at 66°. An isolated occurrence of 10,000 in July is the only record in the summer of 1896. It is in the November-December plankton of 1896 and the March-May plankton of 1897, and attains a pulse of only 6,400 on April 27, at 60°. It does not reappear until the 14th of the following September, in whose stable conditions a pulse of 121,200 on the 21st, at 71°, is found. It disappears October 5, and is irregularly present from January to April, with larger numbers in the latter part of the period. It is not found in 1898 (Table I.) from May 1 to Decem- (14) o o w o o W W m w o < Pi w <^ U5 t-H >-] Q < m W H > Q < o w u p< O W o per cent., are males, 8,542, or 11 per cent., females without eggs, and only 883, or one per cent., egg-bearing females. In so far as these data go, they indicate that this so-called species, or even variety, of C. viridis, in so far as it is based on the lance-like spine of the stylet, is not well founded. This is, it seems, predominantly a male character, though not exclusively so, since females, and even egg-bearing females, are found which exhibit this structure. C. viridis var. brevispinosus appears to be confined to American waters. Marsh ('93 and '95) reports it from the larger lakes of Wisconsin and Michigan, and from the Great Lakes, except Lake Michigan. Birge ('95 and '97) finds that it is the most abundant species of Cyclops (except in summer, when C. leuckarti abounds) in Lake Mendota, and the only one reproducing under the ice. His data exhibit a major pulse in May, and a second one, of less ampli- tude, in October, with slight indications of recurrent minor pulses in midsummer, obscured possibly by the massing of his data in fortnightly averages. The seasonal distribution in Lake Mendota is thus much like that in the Illinois River. Marsh ('97) finds the maximum in Green Lake in June at 68°-69°, and only scattering occurrences at other seasons. E. B. Forbes finds this variety widel}^ distributed in American waters, but never especially abun- dant. Cyclops viridis var. insectus Forbes.—Average number, 539; in 1897, 2,115; in 1896, 949; in 1895, 2,966; and in 1894, 905. It is thus more abundant by two- to threefold in the stable years of 1895 and 1897 than in the flood-swept years of 1896 and 1898. This variety was found in every month of the year, though predominantly in April-October, when the percentages of the collections containing it were respectively 64, 100, 85, 100, 100, 87, and 76 per cent. In November-March the percentages were only 44, 6, 17, 7, and 13. The numbers of individuals are very small, however, from October 1 to April 20, excepting in the autumn of 1897, when, with the delayed high temperatures and the great impetus given to plankton development in the stable conditions of low water, the maximum pulse of all our records, 30,800 per m.^, was reached on October 5, a pulse of 1 ,200 following in November. With 273 these exceptions no record exceeding 600 per m.^* was made between the dates named. Between April 20 and October 1 the minimum records rarely fall below 600 per m.^ except in 1898, and the pulses often culminate at 2,000-8,000. C. viridis var. insectus is thus a planktont of the warmer season, and its seasonal distribution is strikingly similar to that of the so-called var. hrevispinosus. This form occurs in our plankton throughout the whole seasonal range in temperatures, but only in small numbers and irregularly below 60°. Only 21 per cent, of the collections containing insectus were made at temperatures below 60°, and these contained less than 3 per cent, of the total individuals. With the exceptions of the pulses culminating at 43° November 23, 1897, at 1,200 per m.^ and at 57° April 26, 1898, at 4,160 per m.^ no development of this species exceeding 600 per m.^ occurs below 60°. All pulses of more than 3,000 per m.^ excepting only the April pulse of 1898, occur at temperatures above 70°. The species reaches its greatest development in channel waters during the period of maximum temperatures, 70°-80°. The seasonal distribution of this form shows a few straggling individuals in November-March during temperatures below 50°, and a meteoric rise to a vernal pulse in April-May as this tempera- ture is passed and 60°-70° arrives. This is followed by a series of recurrent pulses, often of considerable amplitude, through Septem- ber or until temperatures fall below 60°, as in October, 1897. With falling temperatures the drop in numbers to the winter minimum is quickly accomplished. A comparison of the distribution in 1897 and in other years, shows a close correlation between the decline in temperatures and the falling off in numbers of insectus. The relations which hydrographic conditions bear to the develop- ment of insectus in channel plankton may be inferred from the hydrographs on Plates IX.-XII, Part I., and from the data sum- marized in the following table,—1894 being omitted because of the incompleteness of the seasonal representation. In 1895 levels were low, unusually so in the spring, and the flood-free intervals of the year were of more than the usual extent. About 10 feet of the total movement in levels (51.9 ft.) is found in the late December rise. If this is excluded, the total movement falls to 42 feet, and the range in levels to 6.5 feet. Under conditions, 274 Year 275 Month 276 have their major occurrences in our records as follows, the pulses appearing September 14 and October 5 : July 30 8,080 Sept. 14 117,000 Aug. 10 49,360 Sept. 21 15,260 Aug. 17 17,120 Sept. 29 14,400 Aug. 24.. ...... 20,320 Oct. 5 ...101,600 Aug. 31 67,200 Oct. 12 3,400 vSept. 7 107 , 200 Again, in 1896, the two major pulses of the year are on June 19 (928,984) and July 18 (563,815). Steuer's conclusion seems to be founded upon insufficient data, and can not have general applica- tion. Of the total 240,830 individuals of C. viridis var. insectus in our records in 1894-1899, 117,166, or 49 per cent., are males; 109,460, or 45 per cent., females without eggs; and 14,204, or 6 per cent., females carrying egg-sacs. If the hrevispinosus totals are included, the percentages change to 42 per cent, of females—of which 37 per cent, and 5 per cent., respectively, are without and with egg-sacs — and 58 per cent., males. The apparently high proportion of males may be due to the fact that in the enumeration more 3^oung females than males were included in the "young" Cyclops. The egg-bearing females were generally more numerous in April-July. No marked predominance in the proportion of males appears at any season in our records. Cyclops viridis does not appear extensively in the plankton literature of European lakes. Stenroos ('98) finds it not rare in the littoral fauna of Nurmijarvi See. Scourfield ( '98) reports it as next in abundance to C. semdatus in waters of Epping Forest, where it is perennial. Scott ('99) finds it at all seasons in both littoral and pelagic collections in Scottish lakes, and Amberg ('00) lists it for Katzensee. It appears but infrequently in the investigations of European streams. Neither Schorler ('00) nor Fric and Vavra ('01) report it from the Elbe, though Volk ( '03) lists it from six of seven localities in this stream at Hamburg. Sowinski ('88) finds it in the littoral fauna of the Dnieper, and Zykoff ('03) in the summer plankton of the Volga, though Meissner ( '03) states that it is never found in the plankton of that stream at Saratoff, being confined to the littoral 277 zone and to vegetation. No statistical data concerning its seasonal distribution are given by any of these authors, though Meissner states that it reaches its maximum in I\Iay in the Volga. In addition to the species of Cyclops here listed for the channel plankton of the Illinois, E. B. Forbes ('97) records in May-Septem- ber, 1896, C. varicans Sars as common, and C. fimbriatus var. poppei Rehberg and C. hicolor Sars as rare. Owing to the impossibility of separating with certainty the nauplii and young of the various species of Cyclops they were all recorded together under the head of" nauplii " and " young Cyclops.'' The former includes also the nauplii of the two species of Diaptomus occurring in our plankton. Young Cyclops.—Average number, 4,780; in 1897, 16,035; in 1896, 10,196; in 1895, 21,960; and in 1894, 5,960. With two ex- ceptions in January and February they occur in every collection examined. Numbers are, however, at a minimum in November- March, only 9 instances of more than 1,500 per m.^ appearing in our records in this season. With the exception of two pulses in the autumn of 1897, and two in this season in 1895, all pulses of an amplitude exceeding 8,000 per m.^ are confined to the interval between April 20 and October 1, practically to temperatures above 70°. They also exhibit relations to hydrographic conditions of the same nature as those found in case of the adults of the various species of Cyclops, and manifest likewise the phenomenon of re- current pulses (Table I.). The totals of all young Cyclops in 1894- 1899 are almost five times those of all adults of the genus. This ratio gives an index of the extent of the decimation by enemies and inimical factors of the environment which exists after the nauplius stage has passed and before that of the adult is reached. Nauplii of the Copepoda (excluding the HarpacticidcB) . —Average number, 36,707; in 1897, 53,786; in 1896, 24,560; in 1895, 88,442; and in 1894, 45,648. Nauplii were recorded in all collections ex- amined with but two exceptions. As in the case of the adults and young, the large numbers are, however, confined to the warmer season between April 1 5 and October 1 . During the colder months the pulses rarely rise above 20,000 per m.^, and those in excess of 35,000 during these months are with one exception confined to the delayed high temperatures of the stable autumn of 1897. During 278 the warmer season, on the other hand, the pulses frequently attain 100,000 or over. The maximum record of 928,984 was made in the stable low water of June 19, 1895. All large developments thus lie at tem- peratures above 70°. The nauplii bear much the same relation to hydrographic condi- tions as that found in the adults; for example, in Cyclops viridis. This is seen in the fact that in unstable years such as 1896 and 1898 the numbers are on the average only 28 and 68 per cent, of w^hat they were in the more stable conditions of 1895 and 1897, and the average monthly population in July-December in the unstable conditions of 1898 is only 18 per cent, of that in the same months of the previous year. The relative numbers of adult, young, and larval stages of the CyclopidcB are given in the accompanying table. Year 279 61 in winter conditions of 1899 (January-March), and falling as low. as 1 to 17 in 1896. This was a year of recurrent floods, but its ratio is in sharp contrast with that of 1898 (1 to 30), also a year of con- siderable hydrographic disturbances during the summer. The adult population was reduced during this year, and especially during the summer floods, but the nauplii do not fall conspicuously below those of other years. It would therefore seem that the deleterious action of flood conditions operates more effectively upon the adult and young than upon the nauplii. This fact may be due to the relative absence of spines and hairs on the nauplii, structures which gather silt and load dowm the larger forms in the flood waters. The greater number of young and adults in 1896 as compared with 1898 may be due to the more gradual rise of the floods of the former 3"ear (see PI. X. and XII., Pt. I.) jand the proportionally greater amount of silt in the more sudden floods of the latter. The ratios given in the table are of course subject to the error arising from the uneven seasonal distribution of the collections in some years, and to that arising from varying location of the collec- tions on the pulses, especially on those of greatest amplitude. An ad- ditional error arises from the leakage of the smaller nauplii through the meshes of the silk net. I have found on experiment that they will thus escape under pressure of a column of water only 3-4 cm. in height. Their dimensions are such that the smaller individuals can pass through the meshes of even the No. 20 silk. It seems probable that ratios of nauplii to adults are actually greater than our records indicate. The relationship wdiich the pulses of nauplii bear to those of the adult CydopidcB may be inferred from an examination of the data of Table I. An analysis of the seasonal distribution of the total young and adult CydopidcB and of the nauplii reveals the fact that in all seasons in which collections at approximately weekly intervals were made, their pulses coincide in a majority of cases in their maxima, and when the coincidences do not occur the maximum of the nauplius pulse appears in the collection of the week following that of the young and adult Cydopidce. This appears less constantly and clearly in the disturbed hydrographic conditions of 1898 (Table I.) than in the records of more stable years. Apstein ('96) finds that nauplii of Copepoda are most abundant when eggs are most common, and that this bears no constant relation 280 to the abundance of adults. Our collections, extending over longer periods and being at briefer intervals, indicate, however, that this relation does exist. As above stated, the larvae are most abundant at or shortly after the times of greatest abundance of adults—that is, the maxima of the recurrent pulses. Apstein also states that reproduction is periodic and development rapid. Maximum numbers are reported by him in May and September. Cohn ( '03), on the other hand, maintains that the " innere Logik" and his data show him that the nauplii reach their greatest numbers just prior to the appearance of largest numbers of young and adiilt Copepoda. His data are from 12 collections between May 1 and October 1, and favor his contention in 2 out of 3 cases (of maxima), and both of these lie in collections at intervals of 15 to 16 days. In the light of our data obtained at briefer intervals and the conclusions therefrom that the pulses of larvae tend to coincide or follow at a brief interval those of the adults, it becomes questionable w^hether his data are sufficient for his conclusion. His logic also overlooks the fact, apparently, that smaller numbers of larva: might lead to coincident maxima of grown forms during a period of abundant food, on which all pulses must be based, since the larval stage may he at such times a brief one and the adult a relatively longer one, and the cumulative effect of this relationship would make the conditions shown in our data logically possible. Furthermore, Cohn used a No. 12 silk in his plankton net, and this allows many nauplii to escape, and probably accounts for the fact that the ratio of larvae to grown forms in his figures is only 1.3 to 1, while in our records it is 3 . 5 to 1 . The discrepancy arising from this leakage may further tend to weaken his data for his conclusions concerning the relations of larvae and adults. Steuer ('01) finds that the nauplii in the Danube at Vienna reach maxima in June and in August, but his data are too scattered to fully delineate their fluctuations. Two out of three of his max- ima coincide with those of all Cyclops, and the third antedates it (monthly intervals of collection), as in Cohn's data. Diaptomus pallidus Herrick.—Average number per m.^, 11 ; in 1897, 367; in 1896, 87; in 1895, 152; and in 1894, 146. This species was recorded in all months of the year but February, though in a larger percentage of the collections and in larger numbers in July-December. Prior to this season the percentage does not 281 rise above 31 per cent., the occurrences are irregular, and the num- bers are small. Thus in 1896 and 1898, years of numerous winter and vernal collections, there were but 4 occurrences in each prior to July 1, and all but one of these was of numbers less than 100 per m.^ Only 12 of the 72 occurrences and 8 per cent, of the total individuals were recorded in the first and less stable half of the years. In July-December numbers rise in feebly outlined pulses which attain at the most 800-2,400 per m.^ The percentage of collections con- taining the species rises to 33-75 per cent., and in stable autumns such as 1895 and 1897 the occurrences are but httle interrupted. In its seasonal distribution in channel waters it is thus largely confined to the last—and more stable—half of the year. Its relationship to hydrographic conditions here suggested also appears in a comparison of the yearly averages given above. The average numbers per m.^ in 1896 and 1898, 87 and 11, are greatly exceeded by those of 1895 (152) and 1897 (367). The total number recorded in July-December in 1897 is 29 times that in 1898. This well-defined predominance in stable seasons, which appears also in the case of the closely related D. siciloides, exceeds that of the other Entomostraca, and indicates a greater sensitiveness on the part of these species to the deleterious effects of flood waters. The long antennas and great development of the feathering of the caudal stylets afford a large area for the attachment of the silt and debris of flood waters, and accordingly facilitate the destn.iction or removal of Diaptomus from the plankton more quickly than in the case of Entomostraca in which these processes are less developed—as in Cyclops or Bosmina. The numbers of individuals are too small to delineate accurately the recurrent pulses which are suggested in the data of distribution. In the autumns of 1895 and 1897, when the occurrences are most continuous, the larger numbers tend to fall at the times of the maxima of pulses of other Entomostraca. There is no marked limitation placed upon this species by the seasonal changes in tem- perature. It is found throughout the seasonal range in tempera- tures, though numbers are slightly smaller in channel waters in November-December. Nevertheless it occurs in considerable num- bers in the backwaters in breeding activity under the ice at mini- mum temperatures in December. 282 Of the total individuals, 40 per cent, were males; 45 per cent., females without eggs; and 15 per cent., females with eggs. The sexes show no marked or constant seasonal differences in distribu- tion. Females with eggs are more abundant in August-October, and with spermatophores in the same months. Detached sperma- tophores were found until December. This species is stated by Herrick ('84) to be distributed in the entire Mississippi Valley. Marsh ('93) finds it in Wisconsin, but it appears nowhere in the plankton of the Great Lakes. Brewer ('98) reports it in the backwaters of the Platte in Nebraska, and Schacht ('97) states that it is an exceedingly common species in central Illinois, and that it has been reported from Wisconsin, Ohio, and Minnesota. It thus appears to be limited to the shallow and relatively warm waters of the prairie regions of the Mississippi basin. Diaptonius siciloides Lilljeborg.—Average number, 10; in 1897, 350; in 1896, 56; in 1895, 282; and in 1894, 23. As will be seen on comparison, these yearly averages are very similar to those of the preceding species with the exception that the development of D. siciloides is about twice that of D. pallidus in 1895. In other particulars its seasonal data so resemble those of D. pallidus as to make their discussion in large part a repetition. Its seasonal- distribution relations to temperature and hydrographic conditions, breeding season, and its tendency toward a pulse-like recurrence in coincidence with other Entomostraca are all very similar to these features in D. pallidus. The proportions of the sexes differ slightly, the males being less numerous (3 1 per cent.) and egg-bearing females more abundant (18 per cent.) than in the previous species. This is also an American species, reported thus far only from Lake Tulare, Calif., the Illinois River, and waters of Indiana and Iowa (Schacht, '97), and by Brewer ('98) in lakes and pools of Nebraska. It is thus confined largely to shoal and warm waters. Diaptomus spp., immature.—Average number, 19; in 1897, 560; in 1896, 158; in 1895, 336; and in 1894, 120. The immature individuals of D. pallidus and D. siciloides were not distinguished from each other in the records. Young Diaptomus presumably belonging to these two species occur in every month but March, though but 10 of the 74 records were miade in January- June. The percentage of occurrences and the numbers per m.^ are lowest in these months, not rising above 33 per cent, and 500 per 283 m.^ save in two instances. Occurrences of small numbers continue through July, but from August 1 to October 15 appear the major pulses of the year, attaining an amplitude of 1,000 to 8,800 per m.^ With the decline of temperatures in October, numbers fall to levels below 400 per m.^ with one exception (December 14, 1897) at 700. The percentage of occurrences is, however, high (41 to 44 per cent.) and declines only to 33 per cent, in January. The period of greatest numbers of young thus coincides with that of greatest abundance of adults, and lies at temperatures of 70°, and above, in channel waters. The effect of hydrographic changes upon the occurrence of young Diaptomus appears in striking form in the annual averages above quoted. In 1898, a year of sudden changes, the average per m.Ms only 19, while in the stable conditions of the previous year it is 560. The July-December production in 1897 is 28 times greater than that of 1898. In 1896, a year of recurrent but less sudden floods, the average (158) is less than that of 1895 (336), a more stable year. The great reduction of adults noted in 1898 and 1896 is thus paralleled by an even greater reduction of the young. Osphranticum lahronectum Forbes occurs in the plankton of Quiver Lake in small numbers (see Schacht, '98), and was found once in channel plankton in June, 1896. AMPHIPODA. Allorchestes dentata (Sm.) Faxon.—This is an abundant littoral species found amid vegetation, especially in the vegetation-rich backr^vaters, such as Quiver Lake. It was not often found in channel plankton, being taken only in the summer of 1895, when the July- August floods carried away the vegetation which had accumulated during the antecedent low water. ARACHNIDA. ACARINA. In vegetation-rich backwaters members of the family Hydrach- nidcu were frequently taken, along with other adventitious or- ganisms, with the plankton. In channel waters they are less frequent, and are represented principally by Atax, which is parasitic 284 in great numbers (see Kelly, '99) in the Unionidcs which are found in the bottom of the channel. Occurrences in the plankton were limited to the months of May-August, and may be due in part, especiall}^ in the warmer months, to the release of the parasites by the death and flotation of their hosts. Flood waters in warm months were often disastrous to the UnionidcE because of the load of silt, sewage, and industrial wastes which they carry in channel confines at the lower river stages often prevailing in these months. Other small aquatic A carina were also present, probably adven- titious from the littoral or bottom ooze. With two exceptions their occurrences in the plankton were all in warmer months, April- September, though not in flood waters. During the period of the migration of waterfowl, parasitic Acarina were noted in plankton collections in a few instances. TARDIGRADA. Macrohiotus macronyx Duj.—Average number, 11. This species is found principally in the colder part of the year, from October to May. The earliest autumnal record was October 30, 1895, at 45°, and the latest vernal one. May 1, 1896, at 68.8°, and the maximum number (2,980 per m.^) was recorded on April 10, 1896, at 46.2°. Of this number, one sixth were females with eggs. Females with eggs were also found in November, February, and March. Because of its seasonal distribution it is fotmd principally, though not solely, in disturbed hydrographic conditions, and its occurrence in the plank- ton is largely adventitious. HEXAPODA. Owing to the shoal waters, relatively narrow confines, and the hydrographic fluctuations in our fluviatile environment, the aquatic insects, both larval and adult, have many points of contact with the plankton. They constitute a large element in the total volume of the animal population of shore and bottom, and are all connected by chains of food relations, more or less complex and remote, to the plankton organisms or their sources of food. With the single exception of the larvae of Corethra they are all in the main adventi- tious members of the plankton assemblage, and are much more abundant in the vegetation-rich backwaters than in the channel. 285 Since the aquatic insects of these collections are being studied by- others, with reference to publication in this Bulletin (see Hart, '95, and Needham and Hart. '01), only passing notice of the more important representatives appears in this connection. EPHEMERIDA, Ephemerid larvae, as a rule in early stages, were found singly or in small numbers in the channel plankton in the warmer months, April-October, at temperatures above 56°. Since these occurrences were with few exceptions in stable hydrographic conditions, it seems probable that the younger larvse of this order may adopt, at least temporarily, a limnetic habit. Specific identifications of these larvae were not made. HEMIPTERA. Corisa (?) sp.—Average number, 37. A small hemipterous larva doubtfully referred by Mr. C. A. Hart to Corisa, was taken with some frequency but in relatively small numbers in the plankton during the summer months. Of the 36 occurrences 27 fall in June-August, 2 in May and 3 in September, 2 in January, and 1 each in October and November. It thus appears in the temperature extremes, but exhibits a great predominance in the season of maxi- mum heat. There is no marked increase in its frequency or numbers in years of more disturbed hydrographic conditions. Its numbers are always small and somewhat erratic. Adult Corisa, as well as many other aquatic Hemiptera, were found in plankton collections singly and infrequently. DIPTERA. This group of insects is abundantly represented in the plankton, but in all cases by^ larval or pupal stages. Chironomus spp., larval stages.—Average number, 124. Larvae in various stages of development from that immediately after hatching to that approaching pupation were found in channel plankton. They occur in considerable numbers in the ooze in the river bottom, but appear to abandon the limicolous for the limnetic habit, temporarily at least, as a result of hydrographic or other disturbances. There is evidence from their relative numbers in 286 years of different hydrographic conditions that these have consider- able influence in bringing them into the plankton. Thus in 1897, in stable conditions, there were only 5 occurrences in 3 1 collections examined, averaging 88 per m.^ while in 1898, in more disturbed conditions, there were 29 occurrences in 52 collections, averaging 124 per m.^ There is also a marked seasonal distribution. The larv£e appear in the plankton in March-December through the seasonal extremes of temperature, but the numbers in March and November-December are always small. Only 15 per cent, of the occurrences and 5 per cent, of the individuals were found at tem- peratures below 45°. The percentage of occurrences in the collec- tions is highest in March-September, the percentages being 53, 73, 80, 47, 78, 52, and 50, respectively, to 8 to 35 per cent, during the remaining months. Corethra sp., larval stages.—Average number, 6. These semi- transparent and active larvae have the characteristics of limnetic organisms, and may be reckoned among the autolimnetic planktonts of our waters. Because of their activity , it seems probable that they escape the drawn net,—especially the small model used by us,—and also, because of their negative rheotaxis, elude the suction of the plankton pump to an even greater extent. Thus, in 1895, in net collections, there were 8 occurrences averaging 32 per m.^ to 4 in 1898, in pump collections, averaging 8 per m.^ Corethra larvae were never abundant in our plankton, probably in part for the reasons just cited. With two exceptions all the occurrences lie in the period of maximum temperatures in June-September, 7 of the 14 occurrences and one third of the individuals being recorded in August. Dixa sp., larval stages.—Average number, 8. Larvae were recorded singly in scattered occurrences in all months but Februar}^ and October-December, though most of them appear during maxi- mum temperatures. Larvae of Tanypus and Odontomyia were also recorded in May and June in isolated occurrences. In addition to the larval stages of these aquatic insects there occurred in the plankton a considerable number of insect eggs, principally those of Diptera and Ephemerida. These were generally isolated, though sometimes fragments of the egg-string of Chirono- 287 mils appeared. They were recorded in all months but February and December, though 20 of the 30 records and 81 per cent, of the individuals appeared in IMay-August. The numbers are never very large, the maximum record, 5,424 per m.^ on June 29, 1894, being due to a number of fragments of egg-strings. ]\I O L L U S C A . GASTROPODA. The adults and young of many of our aquatic gastropods have the habit of gliding on the under side of the surface film of water, and they are also frequently dislodged from their foothold on aquatic vegetation, and thus enter the habitat of the plankton temporarily. This is especially true in vegetation-rich backwaters. The smaller forms, such as Ajtcyhis, Amnicola, and Planorhis parvus were occa- sionally taken in the summer plankton of the channel. LAMELLIBRAXCHIATA. This group is represented in the plankton by the larval stages, or glochidia, of the UnionidcB, which form an important part of the bottom fauna of the stream and its tributaries. Anodonta corptdenta Cooper.—Average number of glochidia, 21. The seasonal distribution of the glochidia in the plankton is very well defined. With but two exceptions the 48 occurrences all fall in October-April, and 40 of them in November-March. The occur- rences are thus during the period of minimum temperatures ; indeed, 31 of the 48 are at temperatures not exceeding 35° in surface waters, and only 9 are above 45°. The earliest autumnal record is Septem- ber 30, at 58°, and the latest vernal one, June 6, at 79°. Generally the earliest records are in the closing days of September or the early ones of October, and the latest records are about the first of April. The occurrences are more frequent in December-March, the glo- chidia appearing in 64, 50, 53, and 60 per cent, of the collections, respectively, in these months. Their numbers are also several fold greater at this season than in the earlier and later months of their occurrence. The period of minimum temperatures is thus the season of greatest discharge of glochidia. The numbers are always relatively small, 520 on December 28, 1897, being the maxi- 288 mum record. Their fluctuations are erratic, and show no apparent relation to hydrographic or other environmental changes. Lampsilus anodontoides (Lea) Baker.—Glochidia referred with some uncertainty to this species appeared somewhat irregularly in the plankton in small numbers in September-December and again in June-July. The seasonal distribution in two periods suggests the inclusion of two species. Arcidens confragosus (Say) Simpson.—Glochidia of the type referred by Lea to the old genus Margaritana, and presumably belonging to this the commonest member of this genus (as formerly understood) in our locality, were taken in the plankton December 18, 1895, in small numbers. BRYOZOA. This group is represented in our plankton by the floating stato- blasts, when these occur, as in Pectinatella and Plumatella, by detached and floating fragments, as in Urnatella, or by natant colonies, as in Lophopus and Cristatella. Genera such as Fredericella and Paludicella, whose statoblasts sink, fail to appear in the plank- ton, though in some cases they may be abundant in the bottom fauna. The Bryozoa are plankton feeders, and play an important role as plankton reducers in vegetation-rich backwaters. DISCUSSION OF SPECIES OF BRYOZOA. Cristatella mucedo Cuvier.—This species was found in the back- waters in summer months, especially in Quiver Lake. Statoblasts probably referable to this species occurred sparingly in May and August. Lophopus cristallinus Pallas.—This rare bryozoan occurred in the channel plankton, though not in our quantitative collections, in July, 1897, in that part of the channel containing the discharge from Quiver Lake. Small, free-swimming colonies of 5-50 zooids were taken in surface waters. Pectinatella magnifica Leidy.—Statoblasts of this superb bryo- zoan were not uncommon in the backwaters, and were seen several times in the vernal plankton of the channel. The large floating colonies are found near the surface in July-October in the open backwaters, and more rarely in the river itself. The translucent 289 gelatinous coenoecia are spherical, ellipsoidal, or often somewhat flattened. The longest diameter of these floating masses often exceeds 30 cm. Pliiinatella repens L.—This is by far the most abundant bryozoan in our locality, being found everywhere on submerged vegetation in the backwaters. It often develops with surprising rapidity on the submerged stems of plants, where, as in 1896, summer floods reinvade the vegetation-covered margins of reservoir backwaters. It is represented in the plankton by its floating statoblasts. Their seasonal distribution shows some correlations with temperature, hydrographic conditions, and the seasonal cycle of the parent organisms. During the period of minimum temperatures (Decem- ber-February, inclusive) they are relatively rare in the plankton, appearing in 30, 8, and 20 per cent., respectively, of the plankton catches. They are rare in high- as well as low-w^ater conditions, as, for example, in the floods of 1895-96 and 1898, when they appear in but one of 15 collections. With the rise of temperature in March they occur more frequently, as, for instance, in 1898 (Table I.), and continue during the run-off of the spring flood. The occurrences rise in March-May to 60, 46, and 50 per cent, of our total collections in these months, and the numbers also are larger. For example, in 1898, 81 per cent, of the total individuals for the year were found in these months. The discharge from impounding backwaters, the principal breeding grounds of the parent organisms, doubtless tends to increase the numbers of statoblasts in channel plankton during this season. During the remainder of the year, June-November, the percentage of occurrences again falls to 30, 50, 24, 32, 18, and 44 per cent., respectively. The 50 per cent, in July is due to the summer flood of 1896. If this year is omitted the record falls to 33 per cent. The large percentage for November is probably due to the predominantly higher levels of this month, to the invasion of lake margins seeded with statoblasts, and to the increased activity in the fishing industry, which tends to disturb the summer's growth of vegetation in tributary backwaters. The relations to the seasonal cycle of the species are patent. The summer months, June- September, are the season of growth and spread of the parent organisms and of the formation of statoblasts, especially as receding levels expose the water margins. Hydrographic or other disturb- ances tend to increase the number of statoblasts in the plankton 290 until minimum temperatures are reached, when minimum numbers appear in the plankton. As temperatures rise, the statoblasts tend to float and become more abundant in the plankton, as a result, perhaps, of the physiological and accompanying physical changes in the contents of the statoblast. The declining phase of the major flood of the year is thus the period of greatest flotation and dispersal of the statoblasts. Urnatella gracilis Leidy.—This unique species is found in some abundance on the projecting margins of the shells of the UnionidcB which line the river bottom in many reaches of the channel. Small fragments of the colonies containing only several polypides were found in the plankton in May-August and October. The earliest record was May 25, and the latest, October 25, at 48.5°. The Periodicity in the Multiplication of the Organisms of THE Plankton. One of the most obvious conclusions brought to Hght by the detailed study of the volumetric fluctuations of the plankton pub- lished in Part I. of this report, and most strongly reinforced by the statistical data showing the fluctuations in the numbers of the indi- viduals of the various species and in the sums total of the various biological groups represented in the limnetic fauna and flora, is that plankton production is fundamentally rhythmic or periodic in character, viewed either in its constituent elements or as a whole. This total result is simply the sum of a like phenomenon pervading more or less completely and coincidently the reproductive cycles, the rise and decline in the numbers of the typical constituents of the plankton. The exceptions to this rhythm are usually found in those organisms which are adventitious in the plankton and have their centers of growth and distribution in other regions than the open water. ]\Iany illustrations of this periodic movement in the multiplica- tion of organisms of the plankton have been cited in the preceding pages and may be seen in the accompanying plates. As an illustra- tion for discussion in detail we may take the pulse of July, 1898, shown in the volumetric data of Table III. and Plate XII. of Part I. The fluctuations in the biological population during this period are also tabulated in Table I. of this paper, and graphically presented in Plates II. and IV., which exhibit the movement in the totals of the ChlorophycecE, BacillariacecB, and chlorophyll-bearing Masti- gophora, and of the Rotifera and Crustacea. In the volumetric data the pulse rises from a minimum of .14 cm.^ per m.^ on July 5 to a maximum of .88 cm.^ on the 19th, declining again on the 26th to the second minimum, of .67 cm.^ Its duration is thus four weeks and its amplitude, in comparison with many other pulses in the records, relatively slight. It occurs in the more stable conditions of declining river levels and midsummer tempera- tures. The following list gives the names of the more or less typical planktonts considered in the discussion of this pulse. Others, largely adventitious or insignificant in numbers, might be added (20) 291 292 to the list. Forms whose antecedent minimum does not fall on June 28 or July 5 are designated by a superior 1 ; those whose maximum does not fall on July 19 or 26, by a superior 2 ; and those whose subsequent minimum is not on July 26 or August 2, by a superior 3. The component forms and groups are Crenothrix, etc.\ total SchizophycecB, Microcystis ichthyoblabe^ , total ChlorophycecB, Actinas- trum hantzschii, Crucigenia rectangularis, Pediastrum boryanum^-^-^, P. pertusum^' ^, Raphidium polymorphum^, Scenedesmus genuinus, S. obliquus, S. quadricauda, Schroederia setigera, total BacillariacecE^, Cyclotella kuetzingiana, Diatoma elongatum}, Fragilaria virescens^, Melosira granulata var. spinosa^, M. varians^, Navicula spp., Synedra acus, total ConjtigatcB^, Closterium acerosum, C. gracilis, total Protozoa, total Mastigophora, Eudorina elegans, Euglena acus, E. oxyuris, E. viridis, Glenodinium cinctum, Lepocinclis ovum, Pandorina morum^, Phacus longicauda^' ^' ^, P. pleuronectes^' ^, Platydorina caudata, Pleodorina californica, Trachelomonas acuminata^, T. hispida^, T. volvocina, total Rhizopoda, Diffiugia globulosa, total Ciliata^, Codonella cratera^, Halteria grandinella^' ^, Tintinnidium fiuviatile^' ^, total Rotifera, total Bdelloida^- ^ total Ploima, AnurcBa cochlearis and var. tecta, eggs of A. cochlearis and var. tecta, A. hypelasma, Asplanchna brightwellii^' ^' ^, Brachionus angularis and var. bidens, eggs of B. angularis and var. bidens, B. bakeri and vars. cluniorbicularis^ , melhemi, and tuberculus^' ^ total of all varieties of B. bakeri, B. budapestinensis , B. militarist' ^ B. pala and var. amphiceros, B. urceolaris var. bursarius, B. variabilis"^' ^, Mastigocerca carinata^, Monostyla bulla, Polyarthra platyptera, eggs of P. platyp- tera^'^, Rattulustigris^, Synchcsta pectinata^ , S. stylata, eggsoiSynchceta^ Triarthra terminalis^' ^, Pedalion mirum}' ^, total Entomostraca^' ^ total Cladocera^' ^' ^, Bosmina longirostris^' ^, Ceriodaphnia scitula Chydorus sphcericus,Diaphanosoma brachyurum^,Moina micrura^'^'^ total Copepoda ^' ^' ^, Cyclops viridis var. brevispinosus and var insectus, C. edax, young Cyclops^, nauplii of Copepoda^- ^. An examination of the preceding list and of the qualitative data of Table I . , reveals the fact that 7 1 of the more typical planktonts are found in appreciable numbers in the plankton during this month. To this number we may add 6 immature forms separately listed in the table and 14 group totals, making in all 91 sets of statistical data bearing on the components of this pulse. An analy- 293 sis of the behavior of the constituent species shows that 43 of the 7 1 species (including varieties and forms), 4 of the 6 immature forms, and 10 of the 14 group totals reach their greatest amplitude on the 19th, coincidently with the volumetric maximum. Thus, in all, a total of 57 out of 91, or 53 per cent., of the sets of data are in pre- cise agreement as to the time of maximum development. Fur- thermore, of the remaining 35, there are 10 culminating in the collection prior to the 19th (on the 12th), and 16 on the next subse- quent one (on the 26th,) in all, 26 or 29 per cent, which culminate on immediately contiguous dates of examination. This leaves a residuum of only about 8 per cent, which do not exhibit precise or substantial agreement as to the time of maximum development. In the matter of the location of antecedent and subsequent minima the agreement is less pronounced, possibly because the enumeration error is relatively greater in the case of minimum numbers. We find, however, that 65, or 72 per cent., of the antecedent minima of the pulses occur on June 28 or July 5, and 71, or 79 per cent., of the subsequent minima are on July 26 or August 2. Nineteen, or 20 per cent., of the antecedent minima are on July 12; and 10, or 11 per cent., of the subsequent ones are on August 12. There is thus a residuum of not over 10 per cent, of instances where the data of species or group totals do not coincide or approximate to this pulse, as described, in position of maximum or one or both of the limiting minima. Considering the necessarily large error entering into our data, it is not surprising that exceptions should occur. Some exceptions—as, for example, that of Pediastrum pertusmn (Table I.)—are plainly not due to insufficient data, but are appar- ently normal dislocations ; that is, the rhythm of this species at this time is not in harmony with that of the majority of the components of the plankton. But this is only a temporary derangement, and is not the habitual relationship which movement of production in Pediastrum bears to that of the plankton as a whole. So, also, many of the Entomostraca are much delayed in the culmination of their increase, running over to August 2 or 9, while the most of the other planktonts culminate on July 19 or 26. This lag on the part of the Entomostraca is not, however, habitual, as will be seen on examination of Plates II. and IV. This tendency toward a coincident rhythmic movement in production on the part of the constituent organisms of the plankton will be found throughout all 294 the data where collections are of sufficient frequency to adequately delineate the curve of production, that is from July, 1895, to Oc- tober, 1896, and from July, 1897, to March, 1899, a total of 37 months, and suggestions of a like phenomenon appear in the less complete data of other years. The degree of agreement indicated in the pulse of July, 1898, will be found, on examination of the data in Table I. and in the plates of this paper, to vary with the environ- mental conditions. Times of rapid change in hydrographic condi- tions or in temperature generally show less agreement, and more stable conditions will exhibit an equal or even greater uniformity in the prevalence of the pulse-like rise and decline of the component organisms. In order to show the course of these recurrent pulses in the chlorophyll-bearing planktonts, the total ChlorophycecB, Bacil- lariacecE, and chlorophyll-bearing Mastigophora on the one hand, and of the Rotifera and Entomostraca {"Crustacea'' of the plates), I have presented the data graphically on Plates I.-IV., and in the table on pages 296-299 have drawn up a list of the pulses, indicating the dates of the collections which in the main enter into the respective pulses, and the dates of the maxima or culminations of the five groups named. Owing to the irregularities in the data, there are some instances in which several possible dates miight have been chosen. Reasons for the choice are in several important instances given in the foot-notes to the table. It is evident from the data here presented in graphic and tabular form that the pulses of the five groups of organisms tend in the main to coincide. This is shown in Plates I.-IV., and in the fact that the average divergence of 175 group pulses listed in the table is 6.4 days, or, if 5 aberrant instances are omitted, only 4.8 days. In other words, the pulses of the totals of the 5 groups included in the table culminate on an average within an interval of 6.4 (4.8 in 170 cases) days. The average of the extreme limits between maxima of group pulses in the 36 periods of movement listed in the table is 11.7 days. It is apparent that the pulses would be more completely de- lineated by collections at daily intervals, but even in the somewhat irregular and at times chaotic data here presented, the evidence seems conclusive that the seasonal production of the dominant species and groups of planktonts tends to fall into coincident. 295 recurrent pulses, which, in turn, are the cause of the similar and often coincident volumetric fluctuations. Attention should be directed to the fact that without any im- portant exceptions this recurrent movement pervades all the organisms of the plankton which are eulimnetic,—such as Scenedes- mus, Melosira, Trachelomonas, Codonella, SynchcrM, Daphnia, and Cyclops,—and often those which at certain seasons become tempo- rary planktonts, such as Diffliigia and Hydra, but not with any regularity the tycholimnetic organisms, such as bdelloid rotifers or nematodes. It affects the more highly organized Rotifera and Entomostraca with slow^er growth, longer life, and consequent greater cumulative function as well as the algae, diatoms, and flagellates, where rapid multiplication, brief existence, and non- cumulative (in the individual) function prevail. The large share which the young (eggs and immature stages) play in the pulses of Rotifera and Entomostraca will be seen in Table I., and repeated attention has been called to this in the discussion of species. The prevalence of breeding females and of eggs or young during the rise of the pulse, and of eggless, moribund, or dead individuals or their skeletons during the decline, is a common phenomenon in all well- defined pulses. No species of plankton organisms appears to escape the operation of this recurrent movement in production. The proportion of individuals surviving from one pulse to the next is subject to great variation, being often least when the ampli- tude of the pulses is greatest, and largest when the pulses culminate at slight amplitudes. As a result of periods of minimum develop- ment, it follows that the possible length of life of most plankton organisms, even of the Rotifera and Entomostraca , in the plankton must fall within rather narrow limits of a few days or a fortnight at the most. Since the contrasts between minimum and maximum numbers are relatively greater among the chlorophyll-bearing organisms, it follows that the survival proportion is less in these groups. The duration and amplitude of the plankton pulses will vary within certain limits according to the method of delineation. The volumetric minima and maxima present the total product in cubic centimeters, and the pulses thus marked cat have been described in Part I. They may also be delineated by statistical data of the total plankton or of its larger groups of organisms, or by the domi- w .J o !x O Pi < p hJ o < to CO I-) ^ °° o H ;z; yA 13 iz; O H M ^ O ON o vO o 03 o Q bJj 03 r^ ^ r^ t^ O^ CN "-I --I •r-H 3 X > O bo oJ CN CN (N CN B o3 LTl CN CJ o lO CO CN —< rt o 0) £^ O >. 300 nant or more typical species. In the case of the total plankton some obscurity results at times from the inclusion of unusual proportions of an adventitious population with flood waters. The selection of particular organisms as representative is also subject to some error, since seasonal changes in temperature and other more subtile causes often deflect or suppress their development. The totals of the ChlorophycecE, BacillariacecE, and chlorophyll-bearing Masti- gophora, and of the Rotifera and Entomostraca (PL I.-IV.) probably give as complete and accurate a delineation of the recurrent pulses as the statistical data afford, since they include relatively few adventitious organisms, cover the entire year, and swamp more or less completely individual and temporary divergences of particular species. The delineation of the pulses by statistical data is obviously more significant than the volumetric method, since it more clearly presents the results of the reproductive processes which lie at the foundation of the phenomenon of recurrent pulses ; and this method is also free from the unavoidable error arising from the presence of silt in the collections. The interval between collections introduces an error of consid- erable moment in any effort to determine with accuracy the duration of individual pulses, that is, the length of time between their minima or maxima. Daily collections would render this feasible, but with an interval of a week or more, not only the duration, but in some cases the probable separation of the pulses and location of their maxima, is to some undetermined degree obscured. The duration of the pulses of the five groups of plankton organ- isms shown graphically on Plates I.-Iv., in the case of all chlorophyll- bearing organisms considered as a whole, is in 29 out of 36 instances between 21 and 35 days, less than 21 in 2 cases, and more than 35 in 5, reaching extreme limits of 14 and 49 days. They average 30.25 days between minima and 29.97 between maxima. The rotiferan data in the same months may be divided into 36 periods, in 33 of which pulses are traceable. The duration of pulses between minima lies between 21 and 35 days in 23 of the 36 instances, falls below 21 in 5, and is above 35 in 8. The extreme limits are 14 and 49 days. In the case of the Entomostraca, where also the pulses are obscure in a few of the intervals, we find that 22 of the 36 are between 21 and 35 days between minima, 5 are below 21, and 9 are above 35. 301 The extreme limits are 12 and 49 days, and the average duration is 29.9 days. From the data here presented it is evident that the pulses are in the main from 3 to 5 weeks in duration, averaging approximately 29 + days—a little less than one calendar month. The amplitude of the pulses is affected profoundly by seasonal and local influences, such as the factors of temperature and chemical constituents of the water, and the hydrographic conditions. These have been discussed in connection with the volumetric data in Part I. and in the discussion of species in the first part of the present paper. Rising, or even uniform, temperatures, hydrographic stability, decaying vegetation or access of sewage or other fertilizing constitu- ents, all serve to increase the amplitude of the pulses. Declining temperatures, dilution or suspension of access of fertilizers, competi- tion of gross vegetation, access of flood waters and increase in current, all tend, in the main, to depress the amplitude of the pulses. The duration of the pulses is not, however, thereby essentially modified, though a tendency to override subsequent pulses and partially, rarely wholly, to submerge them is at times of major pulses often apparent in the data. The cause and significance of the phenomenon of recurrent pulses is not clead}^ and unmistakably evident, owing, on the one hand, to the irregularity of the data, and, on the other, to the great complex- ity of the problem, especially in the fluctuations and varying combinations of environmental factors. The plankton method itself is subject to great errors, but these are largely distributed, and careful examination, especially of the matter of dilution and computation, has failed to reveal any probable or even possible source in the method to which these recurrent pulses can be traced. It is not impossible that the rhythm here noted is merely a chance outcome of the statistical method and without biological significance; that it is wholly accidental, the resultant of the con- flicting and varying factors of the environment and not predomi- nantly or continuously initiated by any one factor. On the other hand, its nature, as we have described it, is such that we are led to look for some factor in the environment with which this rhythm of repetition in growth of the plankton organism might be correlated, or to some internal or inherent factor within the organisms constituting 302 the plankton, or to the interaction of environmental and internal factors. That there is a periodicity in the reproductive processes of organisms, of both plants and animals, is generally apparent. We see it in the flowering and fruiting seasons of the phanerogams, and in the breeding seasons of many invertebrates, of mollusks and insects, and of the vertebrates generally,—of fishes, amphibians, reptiles, birds, and most mammals. Fluctuations in environmental conditions, notably in food and temperature, influence these re- productive processes. The phenomenon of rise and decline of the microscopic population in laboratory aquaria is likewise an illustra- tion of the periodicity of organisms, but usually within a briefer interval than thatof the organisms above mentioned. The studies of Maupas ('88) and Calkins ('02) have show^n that even in the seemingly uniform conditions of the laboratory, the reproduction of the ciliate Protozoa is essentially periodic. On a priori grounds it seems highly improbable that in the case of the organisms of the plankton, internal factors should determine the coincidence of the periods of growth and reproduction in several hundred species. While it is not impossible, or indeed improbable, that these species of the plankton if bred in pure cultures or uniform environment would still exhibit a periodic reproduction, it seems highly improbable that so diverse an assemblage of algae, diatoms, flagellates, protozoans, rotifers, and entomostracans as is found in the Illinois River, would exhibit in laboratory cultures under uniform conditions any such coincidence in the location and duration of their pulses as is found in the waters of the stream. Whatever the internal factors involved in the growth and reproduction of plankton organisms may be, it is patent that we must look for some environmental factor or factors lying at the foundation of the coincidence of seasons of growth and reproduction of plankton organisms, which results in the phenomenon of recurrent pulses in species, groups, and volumetric plankton. We may simplify the problem somewhat by recognizing at the outset the importance of nutrition in supplying the basis for the periodic growth of any organism. The rotifers and entomostracans, at least the limnetic types, depend in large measure, either directly or indirectly, upon the synthetic planktonts, such as the algae, dia- toms, and flagellates, for their food. Since the pulses of these animal 303 forms (cf. Plates III. and IV. with I. and II.) coincide with or follow shortly after those of the synthetic planktonts on which they feed, we may conclude that the cause of the periodic movement of these animal groups lies in the periodic fluctuations of their food supply. In the causes which control this periodic growth of the chlorophyll-bearing organisms will be found the solution of the general periodic phenomenon in plankton. This rhythm is primarily one of growth and reproduction, and its solution must be sought in the forms of matter and energy which affect these processes. The nutrition of the chlorophyll-bearing organisms is drawn from matter in the river water. The analyses contained in Part I., Table X., and graphically presented on Plates XLIII. to XLV. trace the seasonal fluctuations in the nitrates—one of the important constituents of plant food. Neither in the seasonal curves of this or other forms of nitrogen delineated in the plates is there any such rhythm of occurrence, though, as has-been pointed out in the discussion of the chemical conditions, there are instances of apparent correlation of plankton and nitrate pulses. They occur at irregular intervals, and do not form a continuous series. That there might be a rhythm in the utihzed nitrates (the analysis repre- sents only the unused residuum) is of course possible, or that it might occur in some other constituent of the food not determined in the analysis is not impossible, but we ha^^e no evidence of its existence. The chlorine in our river w^aters is a fair index of the amount of sewage or pollution by animal wastes. It is subject to considerable fluctuations, resulting in part from dilution by floods or concentra- tion in low waters, and there are other pulses not traceable to hydrographic conditions, which perhaps result from industrial wastes. These fluctuations in some instances coincide with those of the phytoplankton in question, but the instances are few and the correlation is incomplete. Upon investigation I find that sewage pumpage at Bridgeport, which discharged the sewage of Chicago River into the Illinois and Michigan Canal and thence into the Illinois River, w^as practically continuous, and could not produce the rhythm in question. The sewage of Peoria has a much more immediate effect upon the chemical conditions in the river at Havana than has that of Chicago. The sewers of this city, I am informed bv Mr. H. E. Beaslev, Citv Engineer, are flushed as 304 follows: " The method used is that of flushing with a hose, a crew of men being kept constantly at work, taking them about a period of three weeks to cover the entire system. The water is allowed to run through a fire-hose at each point for a period of about ten minutes." This system was in use during the years of our opera- tions, and it offers no occasion for the periodic pulses in growth of the organisms in question. Investigation of the discharges of dis- tillery and cattle-yard wastes into the stream has not revealed any periodic fertilization of the river waters from these sources. The available data thus fail to exhibit any periodic rhythm in food matters in solution and suspension in the river water with which these pulses of chlorophyll-bearing organisms might be correlated. Frequent reference has been made in previous pages to the appearance of pulses upon the decline of floods. Flood waters bring into the river, as shown by the chemical analyses, large quantities of silt and organic wastes in suspension and solution. They inundate great tracts of fertile territory rich in vegetation, and thus add to the available sources of food for the phytoplankton. Decline of the flood affords time for decay and solution of some of the food matters, and time also for breeding, and its run-off adds to the volume of the plankton in channel waters. A comparison of the hydrographs of the years in question (Part I., PL X.-XIII.) with these recurrent pulses (PI. I.) will show that many if not most of the pulses appear on declining flood waters, and that rnany of the larger ones follow the major floods. Closer analysis, however, shows that there are sometimes two pulses of chlorophyll-bearing organisms on the decline of a single flood, and that they may also occur upon rising flood or even in its entire absence. Floods unquestionably affect the amplitude of the pulses, and to some extent modify their location. They seem inadequate, however, to explain their recurrence and their tendency toward a uniform interval. Minima between pulses also recur on declining floods. Energy as well as matter is necessary for the growth of the phytoplankton, and its source is primarily the radiant energy of the sun. A plot of the tri-daily air temperatures at Havana for 1894-1896 (Part I., p. 478, Fig. C) inclusive, exhibits many irregu- larities, a few of which partake of the nature of recurrent pulses at approximately monthly intervals, but they are too few and too irregu- lar to be the basis of the recurrent growth of the phytoplankton. 305 The importance of light for the photosynthesis of chlorophyll - bearing plants is unquestioned. The liberation of oxygen by the plant declines as the light fades, and is at its lowest ebb in darkness. The access of light to the phytoplankton is limited by several factors of the environment, principally by silt, which increases the turbidity, and by clouds, which interfere with the penetration of the sun 's rays. The fluctuations of the silt are chiefly the result of floods, and, as above stated, the floods do not exhibit a rhythmic pulse which can be correlated with that of the phytoplankton; much less do the periods of rising water which are most silt-laden. The cloudiness of the sky varies greatly at different seasons of the year, being predominant at times in the autumn or winter months. It is sub- ject to pulse-like occurrences of variable duration; but an examina- tion of the records for central Illinois for the years under discussion does not disclose any periodic rhythm which can be correlated continuously with that revealed in the statistical records of the growth of the phytoplankton. Another factor of the environment which modifies the quantity of light which impinges upon the chlorophyll-bearing organisms of the plankton is the light from the moon. The amount of light, both absolute and relative, derived from this source is not great. According to the calculations of Zollner, the light from the sun is 618,000 times as bright as that from the full moon. In the pre-' sent connection it is only important to know whether the moon- light contains an amount of solar energy sufficient to appreciably affect the photosynthesis of the phytoplankton. The amount of such energy utilized in photosynthesis is relatively a small propor- tion of the total, so that there is a possibility that moonlight may contribute to the process to an appreciable extent. This matter was investigated by Knauthe ( '98), who determined the fluctuations in the gaseous contents of the waters of carp ponds rich in Euglena. While this author does not report upon the plank- ton of the ponds investigated, it seems quite probable that carp ponds rich in Euglena would present conditions very similar to those found in the Illinois River, which has a remarkably well-developed Euglena water-bloom, and abounds also in carp. The following table presents the results of his work bearing upon the point in discussion. 307 The amount of oxygen present in the water in the dark, or on dark nights, is reported as 0.20, 0.25, and 0.27 cm.^ per 100 cm.^ of water. In bright sunHght in the laboratory, and with the unusual abundance of Euglena due to the collection of the water sample from the region of the water-bloom, it rises to 2.05 cm.^ In the case of the Spandauer samples it rises from 0.25 in the dark to 1.15 (an increase of 0.90 cm.^) after "long" exposure to bright sunlight in the laboratory. The oxygen in this water at 11 :00 p. m., after exposure to moonlight, amounted to 0.45, or 0.20 cm.^ more than was found in control water kept in the dark. In this instance the apparent increase due to moonlight is V9 of that due to sunlight. In the case of the moonlight the analysis was made at 11:00 p. m., after not more than three hours ' exposure. The moon was not at its greatest efficiency, since full moon occurred four days prior to the date of analysis. In the case of the sample exposed to the sunlight the analysis was made at 4:00 p. m., after "langer intens Sonnen- schein." It would seem probable that the effectiveness of moon- light in comparison with sunlight in photosynthesis by the phyto- plankton here indicated (2 to 9) is below the possible maximum and also above that of the average, since it was obtained when the moon was but four days past its maximum effectiveness. If we accept Knauthe 's data as sufficient to establish the effec- tiveness of moonlight in increasing photosynthesis, and thus the growth of the phytoplankton, we find in it a recurrent factor of the environment to whose influence we may seek to attribute the rhythm of growth of the chlorophyll-bearing organisms. On Plates I. and II. I have plotted the seasonal distribution of the totals of the Chlorophycecu, of the BacillariacecB, and of the Mastigophora from July, 1897, to April, 1899, and have indicated the times of full moon throughout this period by marks at the bot- tom of the diagram. The diagram shows clearly the occurrence of these recurrent pulses, their approximation in the three groups of chlorophyll-bearing organisms upon the same or adjacent dates, and the occurrence of their maxima in some cases at the time of full moon or within an interval of ten days thereafter. In the table which follows, I have given the data bearing on the pulses of the total of all chlorophyll-bearing organisms from July, 1895, to October, 1896, and from July, 1897, to March, 1899, inclu- sive, 36 months in all, stating the location of the pulse as determined (21) M ,J U < •z. o w < •J O o I— I Pi < n J; hj >< W o (^ o J K O o w p PL, O o < w -08 > o ^0 < s 3 3 be 3 < < (73 o o > o 1^ 310 in most cases by the delimiting minima, the interval between maxima and that between minima, the date of the maximum, the deviation of the beginning and of the maximum of each pulse from the day of full moon, the deviation of the abscissa of the center of gravity of the polygon formed by the plot of each pulse, and the date of full moon. Deviations prior to the clay of full moon are preceded by the minus sign. The average duration between minima is 30.25 days and that between maxima is 29.97 days; the average location of the initial rise of the pulse is 5.1 days prior to full moon ; and the average lags of the dates of maxima and abscissa of center of gravity of the polygon of occurrences are 11 and 10.45 days, respectively. The probable error of the location of the abscissa of a single pulse is ± 7 . 5 days, and of the average deviation of the abscissa only ± 1.25 days. The table on pages 296-299 shows the lag of the maximum individual pulses of Chlorophycecs, BacillariacecB, chlorophyll- bearing Mastigophora, Rotifera, and Entomostraca. The average lag after the day of full moon for each of the groups, in the order named, is 13.7, 14.8, 14.3, 13.1, and 14.3 days, re- spectively, with a grand average of 14.1 days for the 175 pulses listed. Of these pulses, 135, or 76 per cent., culminate prior to the third week after the date of full moon, and 94, or 52 per cent., in the fortnight between 7 and 2 1 days after full moon. The averages and percentages given in this paragraph vary but slightly from the demands of chance in favor of a hypothesis that the pulses tend to culminate in a particular part of the lunar month, though the data of the total chlorophyll-bearing organisms given above, es- pecially the deviation of the abscissa of center of gravity of the polygon of their occurrences, point in the direction of a lunar factor. There is no doubt of the fact of recurrent pulses and of their distribution at intervals whose average approximates that of the lunar month, though their correlation with any particular part of the month is in no way constant and much less apparent. It would not be strange that the duration interval, or that the position of maxima and minima, should be subject to disturbance, to accelera- tion and dela)^, even to obliteration, in the fluviatile environment with its multitudinous factors,—flood and drouth, summer and winter, clear and turbid waters, bright skies and overcast, the rise and fall of nitrates and other substances in solution or suspension, 311 the fluctuating access of sewage and industrial wastes, the continuous current, the ever-shifting population and the never ceasing struggle for existence and continuance on the part of the interrelated organ- isms of the plankton and of the shores and bottom. The wonder is that any single factor of the environment, however constant, could make any orderly impression in this chaotic situation. This fact that the average interval of the pulses of the phyto- plankton is so nearly the lunar interval would seem to indicate some causal nexus between the two phenomena. An attempt to correlate the plankton pulse with any particular part of the lunar month is, however, less conclusive. The interval of collection, one week, is so great that the course of the pulse can be traced only approximately, since its beginning, maximum, and end can only, from our data, be located at one of these intervals, and more or less distortion results therefrom. Again, the large error in the plankton method may be responsible for some of the fluctuations in the data. Still more potent, probably, are the various factors of the environment of the plankton which combine with the lunar illumination to produce resultants which divert the pulse more or less from the course which the undisturbed lunar factor would cause it to take. Evidence in favor of this view appears in the fact that the greatest disturbances in the rhythmic sequence of the pulses are wont to occur in winter months, when floods, ice, and cloudy weather tend most to interfere with the full action of the lunar factor, while the correlation of full moon and phyto- plankton pulse is most intimate in the stable conditions of summer. This is seen in the fact that the average of the average monthly lags for all of the May-August pulses is 11.9 days, and for the remaining eight months, 18.2 days. The subject here presented is one which lends itself readily to field and laboratory experiment, and it is to be hoped that the sug- gestions of a correlation between the plankton pulses and lunar cycle here made, will be put to the test of further quantitative and statistical, as well as experimental, tests in controlled environments where the disturbing factors of the fluviatile environment are elimi- nated. General Considerations on Seasonal Changes.* It follows from the facts set forth in the preceding discussion that in general each month of the year, characterized by a certain range of hydrographic, thermal, and chemical conditions, and of illumi- nation, has a plankton characterized as follows : — 1. There is a certain range of component species, some of which are occasional stragglers and others more or less uniformly present. 2. There is a certain range of numbers of individuals, varying with the species and profoundly affected by fluctuations in the environmental factors, which change the proportions of the various species from year to year. These proportions vary also from month to month and constitute one of the main elements in the seasonal changes of the plankton. 3. Transitions from month to month are most profound at seasons of greatest environmental change, as, for example, at the times of vernal increase and autumnal decline in temperatures. 4. Seasonal changes in the plankton follow the environmental changes and not the calendar. Autumnal plankton is found when autumnal temperatures arrive. 5. In the main, but two types of plankton are found in the Illinois River—the summer, and the winter assemblage. The vernal and autumnal types are only transitions between the two when organisms from both are present. The winter plankton is charac- terized by a small number of species peculiar to that season, and. a number of perennial forms; the summer, by a larger number of summer organisms with the perennial types. Lake versus River Plankton. Is the plankton of streams (potamoplankton) different from that of lakes (limnoplankton) and ponds (heleoplankton) ? This terminology, introduced by Zacharias ( '98 and '98a), seems to imply a distinction which lies not only in the differences in the configura- * The detailed discussion of seasonal changes in the plankton is deferred to a later paper. 312 313 tion of the basin and in the matter of movement in the water, but also in the constitution of the plankton itself. The examination of the plankton of the Illinois River, and of its backwaters and tributaries, has shown that the plankton of the channel is not im- mediately derived from the tributaries, but comes in large part from the impounding baclavaters, and at low-water stages is almost exclusively indigenous in the channel itself. Upon the basis of the data from the Illinois River the potamoplankton is distinguished from the other types named by the following characters : — 1. It is a polymixic plankton. This is due to the mingling of planktons from all sources in the drainage basin, especially from tributary bacla\'aters, and the consequent seeding of the channel waters with a great range and variety of organisms. In all of our col- lections in channel waters monotonic planktons can scarely be said to be present. The nearest approach to such conditions occurred at low-water stages, when channel waters are most fully isolated. 2. It is subject to extreme fluctuations in quantity and con- stitution. This naturally follows from the manifold factors of the fluviatile environment and the directness with which they impinge upon the plankton. Changes in volume, contact of shore and bottom, access of heat and light, and changes in chemical con- stituents are frequently both more extensive and more widely effective in the stream than they are in the other types of aquatic environment. In consequence, the plankton of the stream is sub- ject to more catastrophic changes than that of the lake. 3. The potamoplankton is not characterized by any species peculiar to it, nor by any precise assemblages of eulimnetic organ- isms. It may be distinguished, in a general way only, by the greater proportion of littoral or benthal forms which are mingled with the more typical planktonts. Zoological Laboratory, University of California, May 10, 1904. i._ 314 TABLE I. Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 r V (o 6B to ^ -a <» - rs ft S ft o o!o Jan. 11.. " 21.. " 25.. Feb. 3 . . 8. . " 15.. " 22.. Mar. 1 . . 8. . " 15.. " 22.. " 29.. Apr. 5 . . " 12.. " 19.. " 26.. May 3 . . " 10.. " 17.. " 24.. " 31.. June 7 . . " 14.. " 21.. - " 28. . July 5 . . " 12.. " 19.. " 26.. 'Aug. 2 . . 9. . " 16.. " 23.. " 30.. Sept. 6 . . " 13.. " 20. . " 27.. Oct. 4.. " 11.. " 18.. " 25.. Nov. 1 . . 8. . " 15.. " 22.. " 29.. Dec. 6.. " 13.. " 20.. " 27.. Average. . 399 105 10 575 275 602 43 79 40 21 18 14 21 21 10 42 7 14 7 10 18 18 43 25 50 10 43 7 21 57 21 25 14 18 28 50 68 34 21 86 1 93 176 190 124 57 136 468 640 600,000 000,000 800,000 000,000 400,000 200,000 200,000 200,000 500,000 600,000 900,000 400,000 600,000 600,000 800,000 000,000 200,000 400 , 000 200,000 800,000 000,000 000 , 000 200,000 200,000 400,000 800,000 200,000 200,000 600,000 600,000 600,000 200,000 400 , 000 000 , 000 000,000 400,000 400,000 200,000 600,000 400,000 800,000 600,000 400,000 800,000 400,000 600,000 800,000 000,000 800,000 497,200 7,477 3,046 14,511 5,440 14 7,200 2,700 9,000 9,000 6,300 1,800 2,701 18,002 190,835 174,000 216,057 50,443 10,800 14,400 21,600 19,800 43,200 18,000 21,600 28,800 162,000 34,200 81,003 1,700,600 212,400 100,817 288,121 118,800 378,013 72,008 111,614 23,400 28,803 3,600 52,201 21,601 10,800 10,400 7,200 7,200 14,400 54,000 59,400 37,800 200 800 837 000 400 800 000 400 600 000 200 100 100 400 200 000 600 200 000 000 000 000 000 000 040 060 400 000 200 000 800 600 600 800 500 000 400 000 500 500 560 000 000 000 000 000 000 000 000 000 200 400 1,600 800 800 500 387 400 800 2,400 800 60 7,200,000 3,000,000 14,400,000 5,400,000 700,000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 9 315 TABLE I— continued. Organisms per Cubic Meter ix Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 ^^ 5 s O' •2* §2 s s s s « a a, s S •2^ a, Jan. 11. " 21. " 25. Feb. 3 . 8. '• 15. •• 22. Mar. 1 . 8. " 15. " 22. " 29. Apr. 5 . " 12. ' 19. " 26. May 3 . " 10. " 17. •' 24. " 31. June 7 . " 14. " 21. " 28. July 5. '• 12. " 19. " 26. Aug. 2 . 9. •' 16. " 23. " 30. Sept. 6. " 13. '• 20. " 27. Oct. 4. " 11. " 18. " 25. Nov. 1 . 8. " 15. " 22. " 29. Dec. 6 . " 13. " 20. " 27. Average . • 1,800,000 1,800,000 10,800,000 5,400,000 3,600,000 5,400,000 5,400,000 10,800,000 21,600,000 10,800,000 7,200,000 7,200,000 1,800,000 1,800,000 900,000 5,400,000 1,800,000 199,038 100 100 3,200 500 c 1,800,000 3,600,000 9 , 000 , 000 10,800,000 3,600,000 900,000 1,800,000 1,800,000 .0 3,000,000 57,600,000 5,400,000 61,200,000 1,800,000 9 , 000 , 000 158,400,000 18,000,000 14,400,000 21,600,000 7,200,000 1,800,000 3,600,000 1,800,000 3,600,000 3,600,000 1,800,000 7,200,000 7,200,000 1,800,000 7,200,000 1,800,000 100 600 600 400 3,200 3,200 6,400 4,800 600 1,000 3,200 32,000 800 6,400 3,600 1,200 4,000 4,000 4,800 2,800 1,600 5,600 8,000 12,000 500 8,500 65,600 5,500 8,000 3,500 18,500 5,000 1,000 3,000 4,000 500 387 300 400 400 1,800 800 200 200 400 1,600 4,800 5,600 1 600 600 12,800 39,400 56,000 55,200 44,800 49,200 136,000 247,600 295,200 145,600 66,400 194,400 326,400 177,600 42 , 000 76,000 259,200 18,500 11,500 9,000 -14,500 3,000 3,000 3,000 2,000 75 640,384 7,153,846 519,231 4,510 j 44,372 316 TABLE I — continued. Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 S-si S «>. (I lo ^ * a s s S « 5. is ^^ ^ a So- lo 2 lO Jan. 11. " 21. " 25. Feb. IS. 22. Mar. 1 . 8. " 15. " 22. " 29. Apr. 5 . " 12. " 19. " 26. May 3. " 10. " 17. " 24. " 31. June 7 . " 14. " 21. " 28. July 5 . " 12. " 19. " 26. Aug. 2 . 9. " 16. " 23. " 30. Sept. 6 . " 13. " 20. " 27. Oct. 4. . 11. . 18. . 25. . Nov. 1 . 8. " 15. " 22. " 29. Dec. 6. " 13. " 20. " 27. 900,000 5,400,000 16,200,000 7,200,000 24,000,000 7,200,000 7,200,000 3,600,000 3,600,000 9,000,000 21,600,000 1,800,000 1,800,000 1,800,000 75,600,000 5,400,000 7,200,000 57,600,000 7,300,000 1,800,000 18,000,000 900,000 7,200,000 10,800,000 5,400,000 1,800,000 3,600,000 3,600,000 1,800,000 1,800,000 900,000 3,600,000 19,800,000 7,200,000 3,600,000 900,000 900,000 900,000 1,800,000 900,000 900,000 10,800,000 1,800,000 36,000,000 1,800,000 900,000 2,700,000 8,100,000 3,600,000 3,600,000 900,000 1,800,000 1,800,000 3,600,000 3,600,000 900,000 900,000 1,800,000 1,800,000 13,500,000 23,400,000 70,200,000 34,200,000 5,400,000 3,600,000 1,800,000 7,200,000 10,800,000 1,800,000 4,500,000 79,200,000 900,000 9,000,000 39,600,000 12,600,000 17,100,000 54,000,000 12,600,000 16,200,000 5,400,000 12,600,000 7,200,000 5,400,000 4,500,000 10,800,000 1,800,000 1,800,000 1,800,000 900,000 14,400,000 6,000,000 7,200,000 9 , 000 , 000 ' 9,000,000 43,200,000 1,800,000 17,100,000 7,200,000 2,700,000 1,800,000 46,800,000 108,000,000 150,000,000 50,400,000 21,600,000 3,600,000 14,400,000 9 , 000 , 000 21,600,000 10,800,000 10,800,000 3,600,000 1,800,000 25,200,000 1,800,000 12,600,000 28,800,000 7,200,000 46,800,000 46,800,000 50,400,000 10,800,000 28,800,000 28,800,000 9,000,000 3,600,000 5,400,000 7,200,000 10,800,000 21,600,000 10,800,000 7,200,000 12,600,000 82,800,000 57,600,000 27,000,000 1,800,000 1,800,000 3,600,000 900,000 7,200,000 3,600,000 1,800,000 1,800,000 900,000 3,600,000 Average . 61,230,769 778,846 1,505,769 9,276,923 21,450,000 519,235 317 TABLE I — continued. Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 •s 5s * a •s « 00 u s^ * s 2 a* Q -i: S an .O 2 W Hi a 00 Jan. 11. ' 21. •• 25. Feb. Mar. July Nov. Dec. 3. 8. 15. 22. 1. 8. IS. 22. 29. Apr. 5 . •' 12. •' 19. •' 26. May 3 . " 10. " 17. " 24. " 31. June 14. 21. 28. 12. " 19. " 26. Aug. 2. 9. " 16. " 23. " 30. Sept. 6. " 13. " 20. " 27. Oct. 4. " U. " 18. " 25. 1. 8. 15. 22. 29. 6. 13. 20. 27. 18 TABLE I— continued. Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries m it are based on filter-paper collections.) 1898 Jan. 11.. ' 21.. " 25.. Feb. 3.. 8. . " 15.. " 22.. Mar. 1 . . 8. . " IS.. " 22.. " 29.. Apr. 5 . . " 12.. " 19.. " 26.. May 3 . . " 10.. " 17.. " 24.. " 31.. June 7 . . " 14.. " 21.. " 28.. July 5 . . " 12.. " 19.. " 26.. Aug. 2 . . 9. . " 16.. " 23.. " 30.. Sept. 6 . . " 13.. " 20.. " 27.. Oct. 4. . " 11.. " 18.. " 25.. Nov. 1 . . 8. . " 15.. " 22.. " 29.. Dec. 6.. " 13.. " 20.. " 27.. Average. . to eg : ^g 1,000 9,090 2,204 3,200 2,800 640 60 30 1 3 2 595 230 3,421 259 109 293 26,028 32,114 153 3,628 1,811 947 133 316 1,484 1,250 366 5,028 1,122 1,200 2,227 5,499 805 840 436 736 83 98 7 60 2 800 240 620 960 800 840 400 440 200 040 360 800 880 120 800 520 520 920 240 000 656 400 800 000 000 000 840 800 000 650 000 200 700 000 000 000 e •fe S t' ^ t: '^ e h 2l> 24,000,000 120,000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 000 14 319 TABLE I — continued. Organisms per Cubic Meter ix Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 320 TABLE I — continued. Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 2 Of St; s-8 at: s 5-S 60 Jan. 321 TABLE I — continued. Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 Jan. 322 TABLE I— continued. Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 323 TABLE I — continued. Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 s a II •S-2 li a. s !§ -8^ Jan. 11 21 •• 25 Feb. 3 8 •' IS " 22 Mar. 1 8 " 15 " 22 " 29 Apr. 5 " 12 " 19 • 26 May 3 " 10 •• 17 " 24 " 31 June 7 " 14 " 21 '• 28 July 5 " 12 " 19 " 26 Aug. 2 9 " 16 " 23 " 30 Sept. 6 " 13 " 20 27 Oct. 4 " 11 " 18 " 25 Nov. 1 8 " 15 " 22 " 29 Dec. 6 '• 13 " 20 " 27 Average > 12,800 835,200 28,800 28,800 28,800 800 3,000 1,600 800 48,000 48 , 400 800 8,000 60,000 40,800 9,600 400 800 12,000 63,200 59,200 1,200 2,400 3,200 2,400 100 500 400 100 400 400 600 200 200 500 400 2,400 8,800 2,000 18,800 49,600 66,800 I 12,000 ,200 ,200 ,400 6,400 1,500 4,800 120,000 900,000 1,200,000 79,200,000 5,400,000 10,800,000 86,400,000 15,300,000 120,000 120,000 1,800,000 180,000 100 600 600 1,600 3,200 3,200 400 200 8,800 8,800 4,800 4,800 3,200 3,200 6,800 800 800 400 400 800 11,200 ; 2,000 4,800 : 8 , 000 ! 4,800! 800 '8,000 1,600 12,800 3,000 7,000 35,200 7 , 000 i 1,500 1,000 500 500 1,000 1,000 1,600 1,000 500 4,800 100 400 400 120 400 60 17,520 ! 6,957 3,711 3,875,769 3,031! 298 17 11 (22) 324 TABLE I — continued. Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 325 TABLE I — continued. Organisms per Cubic Meter ix Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 Jan. 1 1 . " 21. " 25. Feb. Mar. 15. 22. 1. 8. 15. 22. 29. Apr. 5 . . '• 12.. " 19.. •' 26.. May 3. " 10. " 17. '• 24. " 31. June 7. " 14. " 21. " 28. July S . " 12. " 19. " 26. Aug. Sept. Oct. 2. 9. 16. 23. 30. 6. 13. 20. 27. 4. 11. 18. 25. Nov 1. 8. " 15. " 22. " 29. Dec. 6. •' 13. " 20. " 27. § 2 e a! g 5? oCq 5 387 100 1,600 1,400 4,800 4,000 70,400 544,000 29,200 51,200 300,800 6,400 10,400 229,200 272,800 77,600 28,800 80,000 27,000 87,500 409,600 19,000 8,000 8,000 11,500 1,000 100 20 400 43,946 800 200 400 34,800 10,400 93,600 64,800 80,800 138,400 86,400 83,200 10,000 27,500 84,800 9,000 1,000 500 13,973 387 100 1,600 800 200 1,400 4,800 4,000 70,400 544,000 29,600 51,200 335,600 16,800 103,200 292,600 353,600 216,000 115,200 163,200 36,500 115,000 494,400 28,000 9,000 8,500 11,500 1,000 100 20 400 57,919 . • 800 1,600 24,800 128,800 1,600 13,200 72,800 1,200 12,000 105,600 116,000 42,400 28,000 35 331 TABLE I — c o u t i II It c d . Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter^paper collections.) 1898 332 TABLE I — continued. Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 3^ ^ 00 TABLE I — CO n tin ii e d . Organisms per Cubic Meter ix Plaxktox of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 189S Jan. 11. " 21. " 25. Feb. 3. 8. " 15. " 22. Mar. 1 20. 27. _. ^ « •I- 334 TABLE I — continued. Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 fi-p .0 « Q 335 TABLE I — continued. Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 336 TABLE I — continued. Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 337 TABLE I —-continued. Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 338 TABLE I — continued. Organisms per Cubic Meter in Plankton of Illinois River in 1898 (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 339 TABLE I — c o nt i n it c d . Organisms per Cubic Meter in Plankton of Illinois River in 1908. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 Jan. 11 •' 21 " 25 Feb. 3 8 " 15 " 22 Mar. 1 8 " 15 •' 22 " 29 Apr. 5 " 12 " 19 " 26 May 3 " 10 " 17 " 24 ' 31 June 7 " 14 " 21 " 28 July 5 " 12 " 19 " 26 Aug. 2 9 " 16 " 23 •' 30 Sept. 6 " 13 " 20 " 27 Oct. 4 " 11 " 18 " 25 Nov. 1 8 " 15 " 22 " 29 Dec. 6 " 13 " 20 " 27 Average 120 200 770 72 80 160 126 4 800 220 1,120 760 1,420 1,100 5,500 24,320 19,600 53,200 27,900 12,160 5,680 23,800 1 1 , 400 14,600 4,500 1,320 780 1,680 840 360 400 600 3,360 5,520 240 4,320 2,600 3,920 5,580 2,400 300 960 400 160 4 440 1,060 1,800 920 4,780 c •a:a a1 art o c 240 500 2,709 72 80 800 3,159 2 11 16 20 32 25 27 20 182 169 166 126 103 270 164 65 30 10 11 11 2 1 11 13 8 16 19 96 29 27 17 22 14 12 10 6 9 30 640 000 200 800 700 300 100 200 800 400 600 400 400 800 400 800 600 400 400 180 800 600 200 400 600 200 600 000 000 000 000 000 500 500 000 000 000 000 000 000 S 0, a S 000 300 5,400 36,707 40 40 40 20 60 60 40 120 100 12 60 11 S ?i> 40 60 60 60 200 80 40 20 10 a p 900 840 160 80 60 39 0) •v o o'Z. 4,700 774 24 80 400 400 , 600 40 40 40 100 320 4,800 40 40 200 300 60 60 40 120 120 300 40 40 120 120 100 2,000 500 20 318 U J3 O O to 77 100 126 4 40 20 40 40 100 200 200 80 80 200 100 40 180 80 40 60 180 120 160 180 6.0 400 80 40 40 60 180 120 100 320 120 76 S: 40 120 40 80 80 160 300 300 400 400 700 280 440 200 400 400 480 300 480 120 80 60 80 60 200 40 80 120 40 124 (23) 340 TABLE I — concluded . Organisms per Cubic Meter in Plankton of Illinois River in 1898. (An asterisk at head of column indicates that all entries in it are based on filter-paper collections.) 1898 BIBLIOGRAPHY Amberg, C. '00. Beitrage zur Biologic des Katzensees. Inaug. Diss. 78 pp., 5 Taf. Zurich. Also in Vierteljahrsschr. d. naturf. Ges. Zurich, Jahrg. 45, pp. 59-136. Asper, G., und Heuscher, J. '85. Zur Naturgeschichte der Alpenseen. Jahresber. St. Gall. Naturwiss. Ges., 1885-86. 43 pp. Apstein, C. '96. Das Siisswasserplankton. Methode und Resultate der quanti- tativen Untersuchung. Mit 113 Abbildungen. 200 pp., 5 Tab. Kiel und Leipzig. Bertram, [A.] '92. Beitrage zur Kenntniss der Sarcosporidien nebst einem Anhange iiber parasitische Schlauche in der Leibeshohle von Rotatorien. Zool. Jahrb., Abth. f. Anat. u. Ont., Bd. V., pp. 581-604. Birge, E. A. '91. List of Crustacea Cladocera from Madison, Wisconsin. Trans. - Wis. Acad., Vol. VIII., pp. 379-398, PI. 13. '94. A Report on a Collection of Cladocera, mostly from Lake St. Clair, Michigan. Bull. Mich. Fish. Comm., No. 4, pp. 45-49. 1 table. '95. Plankton Studies on Lake Mendota. I. The Vertical Dis- tribution of the Pelagic Crustacea during July, 1894. Trans. Wis. Acad. Sci., Arts, and Letters, Vol. X., pp. 421-484, PI. VII.-X. '97. Plankton Studies on Lake Mendota. II. The Crustacea of the Plankton from July, 1894, to December, 1896. Trans. Wis. Acad. Sci., Arts, and Letters, Vol. XL, pp. 274-448, PI. XV.-XLII. Borge, O. '00. Schwedisches Siisswasserplankton. Botaniska Notiser, 1900, pp. 1-26, Taf. 1. 341 342 Brewer, A. D. "'98. A Study of the Copepoda found in the Vicinity of Lincoln, Nebraska. Journ. Cin. Soc. Nat. Hist., Vol. XIX., pp. 119-138, PI. VII. Brunnthaler, Josef. '00. Plankton-Studien. I. Das Phytoplankton des Donaustromes bei Wien. Verhandl. d. k. k. zool.-botan. Ges. Wien, Jahrg. 1900, Bd. L., pp. 308-311. • '01. Die Coloniebildenden Dinobryon-Arten. (Subgenus Eudino- bryon Lauterborn.) Verhandl. d. k. k. zool.-botan. Ges. Wien. Jahrg. 1901, Bd. LI., pp. 293-306. Mit 5 Fig. im Text. Biitschli, O. '80- '89. Protozoa. Bronn's Klassen und Ordnungen des Thier- reich, Bd. I., Abth. I.-III. 2035 pp., 79 Taf. Leipzig und Heidelberg. Burckhardt, G. '00. Faunistische und systematische Studien liber das Zooplankton der grosseren Seen der Schweiz und ihrer Grenzgebiete. Rev. Suisse de Zoologie, T. VII., pp. 353-713, PI. 19-22. • '00a. Quantitative Studien iiber das Zooplankton des Vierwald- stattersees. Mitteil. d. naturf. Ges. Luzem, 1900, Heft 3, pp. 129-439; Separatabdruck, 311 pp. Burrill, T. J. '02. Report of the University of Illinois. Report of Streams Ex- amination, Chemic and Bacteriologic, of the water between Lake Michigan at Chicago and the Mississippi River at St. Louis, * * * published by authority of the Trustees of the Sanitary District of Chicago, pp. 97-119, PI. XXI.-XXX. (in text). Tab. 1-80. '04. River Pollution and Purification: A Study of the Effect of Chicago Sewage upon the Water Supply of St. Louis. Trans. Am. Micr. Soc, Vol. XXV., pp. 105-120,' Tab. I., PI. XV.-XVII. Calkins, G. N. '02. Studies on the Life-History of Protozoa. I. The Life-Cycle of Paramoecium. caudatum. Arch. f. Entwicklungsmechanik d. Org., Bd. XV., pp. 139-186. 4 fig. in text. 343 Cauller}-, M., et Mesnil, F. '04. Sur les affinites des Actinomyxidies. Compt. Rend. Soc. Biol. Paris, T. 56, pp. 410-412. Cohn, Ludwig. '03. Untersuchungen iiber des Plankton des Lowentin und einiger anderer Seen Masurens. Zeitschr. f. Fischerei, Bd. X., pp. 201- 331, Taf. 1-12. Daday, E. v. '85. Xeue Thierarten aus der Stisswasserfauna von Budapest. Termeszetr. Fiizetek. Vol. IX., pp. 208-215, Taf. XI. Forbes, S. A. '78. The Food of Illinois Fishes. Bull. 111. State Lab. Nat. Hist., Vol. I., No. 2, pp. 71-89. '80. On the Food of Young Fishes. Bull. 111. State Lab. Nat. Hist., Vol. I., No. 3, pp. 66-79; 2d ed., 1903, pp. 71-85. '82. On some Entomostraca of Lake Michigan and Adjacent Waters. Am. Nat., Vol. XVI., pp. 537-542, 640-650, PL VIII., IX. '83. The first Food of the Common Whitefish {Coregonns clupeijormis Mitch.). Bull. 111. State Lab. Nat. Hist., Vol. I., No. 6, pp. 95-109. '90. On some Lake Superior Entomostraca. Rep. U. S. Comm. Fish and Fisheries, 1887, pp. 701-718, PI. I.- IV. '93. A Preliminar}^ Report on the Aquatic Invertebrate Fauna of the Yellowstone National Park, Wyoming, and the Flathead Region of Montana. Bull. L". S. Fish Comm., Vol. XL, pp. 207-258, PI. XXXVII.-XLII. Forbes. E. B. '97. A Contribution to a Knowledge of North American Fresh- water Cyclopidae. Bull. 111. State Lab. Nat. Hist., Vol. V., pp. 27-82, PI. VIII.-XX. Fordyce, C. '00. The Cladocera of Nebraska. Trans. Am. Micr. Soc. Vol. XXII. , pp. 119-174. 4 pi. Also, in 1901, in Studies from the Zool. Lab. Univ. Nebr., No. 42, PI. XXII.-XXV. Fric, A., und Vavra, V. '01. Untersuchungen iiber die Fauna der Gewasser Bohmens. V. Untersuchung des Elbeflusses und seiner Altwasser. Arch, der Naturwiss. Landesdurchforsch. v. Bohmen,' Bd. XL, No. 3, pp. 1-156. Mit 119 Abb. im Text. 344 Friedlander, B. '01. Herrn Alfred Goldsborough Mayer's Entdeckung eines " Atlantischen Palolo" und deren Bedeutung fiir die Frage nach unbekannten Kosmischen Einfliissen auf biologische Vorgange. . Zugleich eine Beleuchtung der darwinistischen Betrachtungsweise. Biol. Centralbl, Bd. XXI., pp. 312-317, 352-366. Fuhrmann, O. '00. Beitrag zur Biologie des Neuenburger vSees. Biol. Centralbl., Bd. XX., pp. 85-96, 120-128. 1 Fig. Garbini, A. '98. Due nuovi Rizopodi limnetici {Dijfiugia cyclotellina— Het- erophrys Pavesii). Zool. Anz., Bd. XXL, pp. 667-670. 2 Fig. Haeckel, E. '90. Plankton Studien. Jenaische Zeitschr., Bd. XXV., pp. 232-336. Hart, C. A. '95. On the Entomology of the Illinois River and Adjacent Waters. Bull. in. State Lab. Nat. Hist., Vol. IV., pp. 149-273, PI. I.-XV. Hempel, A. '96. Descriptions of New Species of Rotifera and Protozoa from the Illinois River and Adjacent Waters. Bull. 111. State Lab. Nat. Hist., Vol. IV., pp. 310-317, PI. XXII.-XXVI. '99. A List of the Protozoa and Rotifera found in the Illinois River and Adjacent Lakes at Havana, 111. Bull. 111. State Lab. Nat. Hist., Vol. v., pp. 301-388. 5 fig. Hensen, V. • '87. Ueber die Bestimmung des Plankton's oder des im Meere treibenden Materials an Pflanzen und Thieren. Fiinft. Ber. d. Komm. zu wiss. Untersuch. d. deutsch. Meere in Kiel, pp. 1-107 + I.-XVIII. 6 Tab. Herrick, C. L. ' '84. A Final Report on the Crustacea of Minnesota included in the Orders Cladocera and Copepoda. Twelfth Ann. Rep. Geol. and Nat. Hist. Surv. Minn., pp. 1-191, PI A-V. 345 Herrick, C. L., and Turner, C. H. '95. Synopsis of the Entomostraca of Minnesota. Geol. and Nat. Hist. Surv. Minn., Zool. Series, II. 337 pp. and 81 pi. Heuscher, J. '85. (See Asper, G., und Heuscher, J.) Huitfeldt-Kaas, H. '98. Plankton in norw^egischen Binnenseen. Biol. Centralbl., Bd. XVIII., pp. 625-636. 2 Fig. Jennings, H. S. '94. The Rotatoria of the Great Lakes and of some of the Inland Lakes of Michigan. Bull. Mich. Fish Comm., No. 3, pp. 1-34, PI. 1 and 2. '96. Report on the Rotatoria. Bull. Mich. Fish Comm., No. 6, pp. 85-93, Fig. 1 and 2. '00. Rotatoria of the United States, with especial Reference to those of the Great Lakes. Bull. U. S. Fish Comm., Vol. XIX., pp. 67-104, PI. 14-22. '00a. A Report of Work on the Protozoa of Lake Erie with especial Reference to the Laws of their Movements. Bull. U. S. Fish Comm., Vol. XIX., pp. 105-114. Jordan, E. O. '00. Some Obsen^ations on the Bacterial Self-Purification of Streams. Joum. Exper. Med., Vol. V., pp. 271-314, PI. XX. Kellicott, D. S. '88. Partial List of Rotifera of Shiawassee River at Corunna, Michigan. Proc. Am. Soc. Microscopists, Vol. X., pp. 84-96. 3 fig. '97. The Rotifera of Sandusky Bay (Second Paper). Proc. Am. Micr. Soc, Vol. XIX., pp. 43-54. Kelly, H. M. '99. A Statistical Study of the Parasites of the Unionidae. Bull. 111. State Lab. Nat. Hist., Vol. V., pp. 399-418. Kertesz, K. '94. Budapest es Kornyekenek Rotatoria-faunaja [The Rotifer fauna of Budapest and Neighborhood]. Inaug. Diss. 55 pp., 1 pi. Budapest, 1894. 346 Knauthe, K. '98. Der Kreislauf der Gase in unseren Gewassern. Biol. Centralbl., Bd. XVIII. , pp. 785-805. Kofoid, C. A.- '95. A Report on the Protozoa observed in Lake Michigan and the Inland Lakes in the Neighborhood of Charlevoix during the Summer of 1894. Bull. Mich. Fish Comm., No. 6, pp. 76-84. '97. Plankton Studies. I. Methods and Apparatus in Use in Plank- ton Investigations at the Biological Experiment Station of the University of Illinois. Bull. 111. State Lab. Nat. Hist., Vol. V.. pp. 1-25, PI. I.-VII. '03. Plankton Studies. IV. The Plankton of the Illinois River, 1894-1899, with Introductory Notes on the Hydrography of the Illinois River and its Basin. Part I. Quantitative Investigations and General Results. Bull. 111. State Lab. Nat. Hist., Vol. VI., pp. vi-xviii, 95-629. 5 text fig. and 50 pi. Kramer, Aug. '97. Ueber den Bau der Korallenriffe und die Planktonvertheilung an den Samoanischer Ktisten. Mit Anhang: Ueber den Palolo- . wurm, von Dr. A. Collin. 174 pp., 1 Karte. Kiel und Leipzig. Lauterborn, R. '93. Ueber Periodicitat in Auftreten und in der Fortpflanzung einiger pelagischer Organismen des Rheins und seiner Altwasser. Verhandl. d. Naturhist.-Med. Vereins Heidelberg, N. F., Bd. V., pp. 103-124. '93a. Beitrage zur Rotatorienfauna des Rheins und seiner Alt- wasser. Zool. Jahrb., Abth. f. Syst., Geogr., u. Biol. d. Thiere, Bd. VII., pp. 254-273, Taf. 11. '94. Ueber die Winterfauna einiger Gewasser der Oberrheinebene. Mit Beschreibungen neuer Protozoen. Biol. Centralbl., Bd. XIV., pp. 390-398. '98. Vorlaufige Mittheilung iiber den Variationskreis von Anurcza cochlearis Gosse. Zool. Anz., Bd. XXL, pp. 597-604. '98a. Ueber die zyklische Fortpflanzung limnetischer Rotatorien. Biol. Centralbl., Bd. XVIII. , pp. 173-183. 347 Leger, L. '04. Considerations sur ie genre Triactinomyxon et les Actinomyx- idies. Compt. Rend. Soc. Biol. Paris, T. 56, pp. 846-848. Leidy, Joseph. '79. Fresh-Water Rhizopods of North America. Rep. U. S. Geol. Sui-v. Terr., Vol. XII. 324 pp., 48 pi. '87. Asplanchna ebbesbornii. Proc. Acad. Nat. Sci. Phila., 1887, p. 157. Lemmermann. E. '00. Beitrage zur Kenntniss der Planktonalgen. XI. Die Gattung Dinobr^-on Ehrenb. Ber. d. deutsch. Bot. Ges., Bd. XVIII., pp. 500-524, Taf. XVIII., XIX. Levander, K. M. '94. Kleine Beitrage zur Kenntniss des Thierlebens unter dicker Eisdecke in einiger Gewassem Finlands. Meddel. Soc. pro Fauna et Flora Fennica, Bd. XX., pp. 66-71. '00. Zur Kenntniss der Fauna und Flora finnischer Binnenseen. Acta Soc. pro Fauna et Flora Fennica, Vol. XIX., No. 2, pp. 1-55. 1 Fig. Marsh, C. Dwight. '93. On the Cyclopidas and Calanidae of Central Wisconsin. Trans. Wis. Acad. Sci., Arts, and Letters, Vol. IX., pp. 189-224, PI. III.-VI. '95. On the Cyclopidae and Calanidae of Lake St. Clair, Lake Michi- gan, and certain of the Inland Lakes of Michigan. Bull. Mich. Fish Comm., No. 5. 24 pp., 5 pi. '97. On the Limnetic Crustacea of Green Lake. Trans. Wis. Acad. Sci., Arts, and Letters, Vol. XL, pp. 179-224, PI. V.-XIV. Marsson, M. '00. Zur Kenntnis der Planktonverhaltnisse einiger Gewasser der Umgebung von Berlin. Forschungsber. aus der Biol. Station zu Plon, Teil 8, pp. 86-119. Haskell, W. M. '87. On the Fresh-water Infusoria of the Wellington District. Trans. New Zealand Inst., Vol. XIX., pp. 49-61, PL III.-V. 348 Maupas, E. '88. Recherches experimentales sur la multiplication des Infusoires cilies. Arch, de Zool. Exper. (2), T. VI., pp. 165-277. 4 pi. Mayer, A. G. '00. An Atlantic "Palolo," Staurocephalus gregaricus. Bull. Mus. Comp. Zool., Vol. XXXVI., pp. 1-14, PI. I.-III. Meissner, V. '02. Notiz uber niedere Crustaceen des Wolga-Flusses bei Saratow. Zool. Anz., Bd. XXVI., pp. 51-55. '03. Contribution to the Fauna of the lower Crustacea of the River Volga. Year-book of the Volga Biol. Station, 1903. 43 pp., 2 tab., PI. II.-IV. (Russian.) Minkiewitsch, R. '98. Some Observations on the Protozoa, made at the Biological Station at Bologoe. Rep. Imp. Soc. Nat. St. Petersburg, Vol. XXIX., pp. 239-246, 267-269. Mrazek, A. '00. [Review of Stole, '99.] Zool. Centralbl., Bd. VII., pp. 594-595. Mit 2 Fig. im Text. Needham, J. G., and Hart, C. A. '01. The Dragon-flies (Odonata) of Illinois, with Descriptions of the Immature Stages. Part I. Petaluridse, ^schnidae, and Gomphi- dse. Bull. 111. State Lab. Nat. Hist., Vol. VI., pp. 1-94, PI. I. Ostwald, Wolfgang. '03. Theoretische Planktonstudien. Zool. Jahrb., Abth. f. Syst., Geogr., u. Biol. d. Thiere, Bd. XVIII., pp. 1-62. Mit 4 Curven und 2 Abb. im Text. '03a. Ueber eine neue theoretische Betrachtungsweise in der Plank- tologie. Forschungsber. a. d. Biol. Station zu Plon, Teil X., pp. 1-49. Penard, E. '02. Faune Rhizopodique du Bassin du Leman. Avec nombreuses figures dans le texte. 714 pp. Geneve. Plate, Ludwig. '85. Beitrage zur Naturgeschichte der Rotatorien. Jenaische Zeitschr., Bd. XIX., pp. 1-120, Taf. I.-III. 349 Prowazek, S. '00. Synedra hyalina, eine apochlorotische Bacillarie. Oesterr. botan. Zeitschr., Bd. 50, pp. 69-73. 2 Fig. Przesmycki, A. M. '01. Ueber parasitische Protozoen aus dem Inneren der Rotatorien. Anz. Akad. Wiss. Krakow, 1901, pp. 358-408. 3 Taf. Ross, L. S. '97. A New Species of Daphnia, and Brief Notes on other Cladocera of Iowa. Proc. Iowa Acad. Sci., Vol. IV., pp. 162-166. 1 fig. Rossinski, D. M. '92. Materialien zur Kenntniss der Evertebratenfauna des Moskwa- fltisses. Bull. Soc. Hist. Nat., Anthrop., Ethnogr., T. LXVIL, Stippl. a Nr. 6, pp. 1-38. Rousselet, C. F. '97. Brachionus Bakeri and its Varieties. Joum. Quekett Micr. Club., Ser. II., Vol. VI., pp. 328-332, PI. XVI. '02. Third List of New Rotifers since 1889. Joum. R. Micr. Soc, 1902, pp. 148-154. '02a. The Genus Synchaeta: A Monographic Study with Descrip- tions of five New Species. Journ. R. Micr. Soc, 1902, pp. 269-290, 393-411, PI. III. -VIII. Roux, J. '99. Obser\-ations sur quelques Infusoires Cilies des Environs de Geneve, avec la description de nouvelles especes. Rev. Suisse de Zool., T. 6, pp. 555-636, PI. XIII, XIV. '01. Faune Infusorienne des Eaux stagnantes des Environs de Geneve. Mem. Inst. Nat. Geneve, T. XIX. 148 pp., 8 pi. Sand, R. '01. Etude monographique sur le groupe des Infusoires Tenta- culiferes. 441 pp., 24 pi. Bruxelles, 1901. Published also in Ann. de la Soc. Beige de Micr., T. XXIV., pp. 57-189; T. XXV., pp. 7-205; T. XXVI., pp. 11-119. Say, T. '18. An Account of the Crustacea of the United States. Journ. Acad. Nat. Sci. Phila., Vol. I., pp. 423-458. 350 Schacht, F. W. '97. The North American Species of Diaptomus. Bull. 111. State Lab. Nat. Hist., Vol. V., pp. 97-207, PI. XXI.-XXXV. '98. The North American Centropagidae belonging to the Genera Osphranticum, Limnocalanus, and Epischura. Bull. 111. State Lab. Nat. Hist., Vol. V., pp. 225-269. Schewiakoff, Wl. '93. Uber die geographische Verbreitung der Siisswasser-Protozoen. Mem. I'Acad. Imper. des Sci. de St.-Petersbourg, Ser. VII., T. XLL, No. 8. 201 pp., 4 Taf. Schorler, B. '00. Das Plankton der Elbe bei Dresden. Zeitschr. f. Gewasser- kunde, Bd. III., pp. 1-27. Schroder, B. '97. Attheya, Rhizosolenia und andere Planktonorganismen in Teiche des botanischen Gartens zu Breslau. Ber. d. deutsch. botan. Ges., Bd. XV., pp. 367-374, Taf. XVII. '97a. Ueber das Plankton der Oder. Ber. d. deutsch. botan. Ges., Bd. XV., pp. 482-492, Taf. XXV. Scott, Thomas. '99. The Invertebrate Fauna of the Inland Waters of Scotland. Report on Special Investigation. Ann. Rep. Fish. Board Scot- land, Vol. XVII., pp. 132-204, PI. VII. Scourfield, D. J. '98. The Entomostraca of Epping Forest, with some General Re- marks on the Group. Essex Naturalist, Vol. X., pp. 193-210, 259-274, 313-334. Seligo, A. '00. Untersuchungen in den Stuhmer Seen. Nebst einemi Anhange: Das Pflanzenplankton preussischer Seen, von B. Schroder. 88 pp., 9 Tab., und 10 Taf. Danzig. Sharpe, R. W. '97. Contribution to a Knowledge of the North American Fresh- water Ostracoda included in the Families Cytheridas and Cyprid- idse. Bull. 111. State Lab. Nat. Hist., Vol. IV., pp. 414-482, PI. XXXIX.-XLVIII. 351 Skorikow, A. S. '96. Rotifera of the Neighborhood of Charkow. Contrib. Soc. Nat. Imp. Univ. Charkow, Vol. XXX., pp. 209-374, PI. VI.-IX. (Russian.) '97. Contribution to an Investigation of the Distribution of the Invertebrates of the River Udy. Contrib. Soc. Nat. Imp. Univ. Charkow, Vol. XXXI., pp. 39-48. (Russian.) '02. Die Erforschung des Potamoplanktons in Russland. Biol. Centralbl., Bd. XXII., pp. 551-570. Smith, F. '94. List of the Protozoa and Mollusca observed in Lake St. Clair in the Summer of 1893. Bull. Mich. Fish Comm., No. 4, pp. 42-44. '00. Notes on Species of North American Oligochaeta. III. List of Species found in Illinois, and Descriptions of Illinois Tubificidae. Bull. 111. State Lab. Nat. Hist., Vol. V., pp. 441-458, PI. XXXIX., XL. Smith, S. I. '74. Sketch of the Invertebrate Fauna of Lake Superior. Rep. U. S. Fish Comm., 1872-73, Part II., pp. 690-707. Sowinski, W. K. '88. Beschreibung der Fauna der Slisswassercrustaceen aus der Umgebung Kiews und des nordlichen Teiles des Gouvernement Kiew. Schriften der Kiew 'schen-Naturf . Ges., 1888, Bd. IX., pp. 225-298. Mit Karte. Stenroos, K. E. '97. Zur Kenntnis der Crustaceen-Fauna von Russisch-Karelien. Cladocera, Calanidae. Acta Soc. pro Fauna et Flora Fennica, Bd. XV., No. 2. 72 pp. Mit einer Tafel und einer Karte. '98. Das Thierleben im Nurmijarvi-See. Acta Soc. pro Fauna et Flora Fennica, Bd. XVII., pp. 1-259, Taf. I-III. 1 Karte. Steuer, Adolf. '01. Die Entomostrakenfauna der "alten Donau" bei Wien. Eine ethologische Studie. Mit einem Anhang: Zur Frage liber Ursprung und Verbreitung der Entomostrakenfauna des Siisswas- sers. Zool. Jahrb., Abth. f. Syst., Geogr., u. Biol. d. Thiere, Bd. : XV., pp. 1-156, Taf. I.-XII. Mit 20 Abb. im Text. 352 Stingelin, Th. '97. Ueber jahreszeitliche, individuelle und locale Variation bei Crustaceen, nebst einigen Bemerkungen iiber die Fortpflanzung bei Daphniden und Lynceiden. Forschungsber. a. d. Biol. Station zu Plon, Th. V., pp. 150-165. Mit 5 Fig. im Text. Stokes, A. C. '86- '88. A Preliminary Contribution toward a History of the Fresh- water Infusoria of the United States. Journ. Trenton Nat. Hist. Soc., Vol. I., pp. 71-319, PI. I.-XIII. '94. Notices of presumably undescribed Infusoria. Proc. Am. Phil_ Soc, Vol. XXXIII., p. 338-345, PI. XXI. Stole, A. '99. Actinomyxidia, nova Mesozou, pfibuzna Myxosporidiim. Rozp. Geske Ak. Cis. Frant. Jos. II., VIII., 22. 12 pp., 3 pi. Thorpe, V. G. '93. The Rotifera of China. Journ. R. Micr. Soc, 1893, pp. 145- 152. 2 pi. Verworn, M. '97. Die polare Erregung der lebendigen Substanz durch den con- stanten Strom. Archiv f. Physiologic (Pfiiiger), Bd. 65, pp. 63-81, Taf. IV. Volk, R. '03. Hamburgische Elb-Untersuchung. I. Allgemeines iiber die biologischen Verhaltnisse der Elbe bei Hamburg und iiber die Einwirkung der Sielwasser auf die Organismen des Stromes. Mitth. a. d. Naturhist. Mus. in Hamburg, Bd. XIX., pp. 65-164, Taf. I.-VII. Ward, H. B. '96. A Biological Examination of Lake Michigan in the Traverse Bay Region. Bull. Mich. Fish Comm., No. 6. 100 pp., 5 pi. Weber, E. F. '98. Faune Rotatorienne du Bassin du Leman (2™<^ partie). Rev. Suisse de ZooL, T. V., pp. 355-785, PL 16-25. Weismann, A. '79. Beitrage zur Naturgeschichte der Daphnoiden. VI. Samen und Begattung]|der Daphnoiden. VII. Die Enstehung der cyclischen Fortpflanzung bei den Daphnoiden. Zeitschr. f. wiss. ZooL, Bd. XXXIII. , pp. 55-270, Taf. VIII.-XIII. 353 Wesenberg-Lund, C. '98. Uber danische Rotiferen und uber die Fortpflanzungsverhalt- nisse der Rotiferen. Zool. Anz., Bd. XXL, pp. 200-211. 'GO. Von dem Abhangigkeitsverhaltnis zwischen dem Bau der Planktonorganismen und dem spezifischen Gewicht des Slisswas- sers. Biol. Centralbl., Bd. XX., Nr. 18 und 19, pp. 606-619, 644-656. Whipple, G. C. '94. Some Observations on the Growth of Diatoms in Surface Wa- ters. Tech. Quart., Vol. VII., pp. 214-231. Whipple, G. C, and Jackson, D. D. '99. Asterionella, its Biology, its Chemistry, and'its Effect on Water Supplies. Journ. X. E. Waterworks Ass'n, Vol. XIV., pp. 1-25. 4 pi. Woodworth, W. McM. '97. On some Turbellaria from Illinois. Bull. Mus. Comp. Zool., Vol. XXXI., pp. 1-16. 1 pi. Zacharias, O. '94. Faunistische Mittheilungen. Forschungsber. a. d. Biol. Station zu Plon, Th. II., pp. 57-90, Taf. I., II. Mit 2 Fig. im Text. '97. Neue Beitrage zur Kenntniss des Stisswasserplanktons. For- schungsber. a. d. Biol. Station zu Plon, Th. V., pp. 1-9, Taf. I. *97a. Die Fauna der Versuchgteiche. Zeit. f. Fischerei, Bd. V., pp. 37-51. '98. Das Heleoplankton. Zool. Anz., Bd. XXL, pp. 24-32. '98a. Das Potamoplankton. Zool. Anz., Bd. XXL, pp. 41-48. '98b. Untersuchungen uber das Plankton der Teichgewasser. For- schungsber. a. d. Biol. Station zu Plon, Teil VI., pp. 89-139, Taf. IV. '00, Trichodina pediculus Ehrb. als Mitglied des Planktons der Bin- nenseen, Biol. Centralbl., Bd. XX., p. 463. Zemow, S. A, '01. Remarks on the Zooplankton of the Rivers Schoschma and Wjatka (Region Malmyz, Government Wjatka). Bull. Imp. Soc. Friends Sci. Moscow, Zool. Section, Vol. 3, pp. 28-35, 36. 1 pi. 354 Zimmer, C, und Schroder, B. '99. Das Plankton des Oderstromes. Forschungsber. a. d. Biol. Station zu Plon, Teil VII., pp. 1-24. Zumstein, H. '99. Zur Morphologic und Physiologic der Euglena gracilis Klcbs. Inaug. Diss. 50 pp., 1 Taf. Leipzig. Also, in 1900, in Jahrb. f. wiss. Botanik, Bd. XXXIV., pp. 149-198, Taf. VI. Zykoff, W. '00. Das Potamoplankton der Wolga bci Saratow. Zool. Anz., Bd. XXIII., pp. 625-627. '03. Report on the Fauna of the Volga and the Aquatic Fauna of the Government of Saratoff. Bull. Soc. Naturalists of Moscow, 1903, pp. 1-148, PL I., II. (Russian.) EXPLANATION OF PLATES. Plate I. Seasonal distribution of synthetic groups of planktonts, Chlorophycece , Bacillariacea:. and Alastigophora, from July 1, 1895, to October 1, 1896. Note changes of scale indicated at bottom of diagram. Numbers in column at left apply only to 1895. In this plate and in II. and IV., apices exceeding the limit of the diagram are dropped down between dotted lines to show location. Circles at bottom indicate location of day of full moon. Plate II. The same as above, from July 1, 1897, to April 1, 1899. Note change in scale from previous plate. Plate III. Seasonal distribution of total Rotifera and Crustacea from July 1, 1895, to October 1, 1896. The Crustacea included, belong almost exclusively to the Ento- mostraca. Apices exceeding the limits of the diagram are dropped down between dotted lines to show location. Totals include both adult and immature stages of the Entomostraca when detached from parent, and both free and attached eggs of the Rotifera. Plate IV. The same as above, from July 1, 1897, to April 1, 1899. Plate V. Seasonal distribution of Polyarthra platyptera. Total number of individuals, not including eggs, represented by ordinants, parts of which exceeding 200,000 are represented by diagonal lines instead of solid vertical lines. Thus parts of a seasonal plot which overlap those above it on the plate are represented by the diagonally-lined ordinants. (24) 355 356 ^ ^ 357 —; ii_L 358 ^LD 359 -. 1 il ' i . 1 [ : ' 1 M k; . 360 J_ ERRATA AND ADDENDA. Page 58, line 7, for ovalis read ovata. Page 85, line 8, for longicaudus read longicaiida. and just above Phaciis pleuro- nectes read the following paragraph : — Phacus longicaiida var. torta, n. var.—This variety, for which I propose the name torta because of the twisted body, is figured by Stein ('78, Taf. 20, Fig. 3). It occurred sparingly in midsummer froin July to September, rarely in October, in 1896 and 1897. Page 91, line 18, after T. catidata Ehrb. read T. lagenella Stein. Pages]153, line 3 from'bottom, 168, line 16, and 178, line 14, for '98 read '98a. Pages 156, line 11, 159, line 16, and 161, line 5 from bottom, for '93 read '98a. 361