Bulletin OF THK Illinois State Laboratory Olf Natural History Urbana, Illinois, U. S. A. STEPHEN A. FORBES, Ph.D., LL,.D. DiRBCTOR Vol. IX. August, 1913 Article XI. VEGETATION OF SKOKIE MARSH BY Earl E. Shekkk ERRATA AND ADDENDA Page 54, lines 3 and 2 from bottom, and elsewhere in Article III. for Cassia chainaechrista read Cassia chamaccrista. Page 62, between lines 4 and 5 from bottom of table insert Erigeron annuus. Page loi, table, after Croloii glandulosus read var. septentrionalis: and for Eijuisettim laeinyatum read Eqnisetum bycmale var. inlermedium. Page 131. line 3, for coerulea read caerulca. Page 138, last line, for Zi::a read Zizia. Page 141, line 21 from bottom, dele Diodia teres. Page l6g. between lines 3 and 4, insert as follows : Erigeron annuus (L.) Pers. An interstitial in the liunch-grass association in the Hanover area. Page 177, line 5, for casti^'ard read iveslzvard. Page 209. line 3 from bottom, for copalina read copallina. Page 210, line 13 from bottom, for Diospyrus read Diospyro'i. Page 211, line 5, for Foresteria read Foresticra. Page 256, line 3 of table, for Dr. H. M. Pepoon read H. 5". Pepoon. Page 278, line 16, the fifth word should be in Roman type. Page 286, line 6 (second column), page 295, list of secondary species (second column), and page 353. line 8 from bottom, for hiematis or hiemale read hye- mnte. Page 313, line 4 from bottom (first column), for pedicularis read pcdicularia. Page 315, line 10, second column, for Apoeynum read Apocynum. Page 323, line 3 from bottom, for Cyperus read Scirpus. Page 330, line 14, for virginianum read inrginicum. Page 336. lines 3 and 2 from bottom, for virginicum read virginianum. Page 337. line 2 from bottom, for philadelphicum read philadelphicns. Page 339» in first list of invading species, for Rhus hiria read Rhus typhiua. Page 351. line 4 from bottom, for .verophtic read .verophytic. Page 355, above line 6 from bottom, insert Scirpus heterochaetus Chase. Page 356, line 14 from bottom, for Symlocarpus read Symplocarpus. Page 360, line 14, for Pirus read Pyrus. Page 362, after line 7, insert Acer saccharinum L. Page 363, line 2 from bottom, for quadiflorum read quadriflorum. Page 365, line 14, for Ihapus read thapsus. Page 369, last line, for Tanecelum read Tanaceium. Page 417, line i. dele the. Page 497, line 9 from bottom, for neglible read negligible, and in foot-note, for Auslall read Anstalt. Page 498, line 4 from bottom, for Lockport read Chillicothe. Page 500, line 13 from bottom, after up insert in. Page 5or, line 2 from bottom, for dissolving read dissolved. Page 504, line 23, for gryina read gyrina : line 17, for dentata read knickerbockeri. Page 506, line 11, for vernata read ternala. Page 507, line 3 from bottom, for Macon read tt'asoi. Page 513, line 19, for Nepa read Zaitha; line 18, and page 517, line 13 from bot- tom, page 520, line 12 from bottom, and page 532, line 4, read naid or naids for naiid or naiids. Page 517, line 6 from bottom, for pondiveed read pickcrel-iveed. Page 519. for first sentence of last paragraph read as follows: We have no exactly comparable chemical data for July; but analyses for August give percentages of saturation for Morris and Marseilles as follows : 20.4 per cent, at Morris on the nth and 11 per cent, at Marseilles on the 12th; 16.35 P<^r cent, at Morris on the 22d and 23d and 7.4 per cent, at Marseilles on the 24th and 25th. Page 521, line 6 from bottom, and page 529, line g, for chrysoleucas read cryso- leticas. Page 525, line 22, and page 536, lines 21 and 24, for Ekmann read Ekmnn. Page 532, line i, for Ancyclus read Ancylus. Page 551, line 7, for 00 read 572. Page 615, second line above foot-note, for 106 read 94. Page 616, line I, for the second Biindeln read Bilndel; line 2, for Biindeln read Biindcls; line 3, for ausscrn read ausscren; line 6, for sweierlie read sweierlei. Page 629, line 12, for kciii read kcincn. Page 634, line 9, for untcrnommcn read untcninntincnen; and in line 14 from bottom, after ;;/$ insert is fig. Plate III, Fig. i, after the word mixed in legend insert consocies of the. Plate IX, Fig. 2, dele the legend and read instead ; Root-system of Tephrosia virgiiiiana, exposed by blowing of the sand. Plate X, Fig. 2, dele the legend and read instead ; A blowout almost stabilized by bunch-grasses, especially Leptoloma cognatum. Plate XXXIX. for Calainogrostis read Calamagrostis. Plate LIV. exchange places of cuts, but not the legends. Plate LXXXV, for 7 read ye. Article XI. — Vegetation of Skokie Marsh. Bv Earl E. SherFF. With the rapid encroachment of city and town upon the outlying districts about Chicago, and the consequent despoilation of the native flora, it has seemed to the writer advisable to undertake a careful study of a certain restricted area, while there is yet an opportunity, and to place these results on record. For several reasons, Skokie IMarsh was deemed most worthy of study. During the past few years the so-called "North Shore" towns situated in the vicinity of the marsh have grown at a phenomenal rate. Much land but re- cently used in farming is now occupied by residences. Moreover, wuth further increases in population it appears certain that the whole marsh area will be thoroughly drained and, as a result, its floristic complexion be entirely changed. However, at the present time the flora is still essentially virgin in many places, and it is reasonably sure that the general survey here presented approximates closely to a truthful statement of natural conditions. The general features of the flora and topography were studied mainly in the autumn of 1910 and the spring of 191 1. From May to October, 191 1, rather intensive taxonomic and ecological studies of the flora were pursued. Again, in 1912, frequent trips were made through various parts of the marsh to secure additional information as a check upon that already obtained. Numerous specimens of plants were gathered from time to time. Of these a considerable number are now in my private herbarium; and many duplicates are in the Herbarium of the Field Museum (Chi- cago), the ^Missouri Botanical Garden Herliarium (St. Louis), Gray Herbarium of Harvard University (Cambridge), the United States National Herbarium (Washington), and the Herbarium of the Royal Botanic Garden (Edinburgh). The data secured, and here published for the first time in collected form, have already appeared in part in several other publications, which are cited in the appended list of literature. The map (PI. LXXXVI, Fig. i) is intended to portrav merely the general location and extent of Skokie Marsh; hence cer- tain of the roads running across the marsh are omitted. All the illustrations were made by the writer, resort being had to pen sketches where photographs were found impracticable. Grateful acknowledgment is here made of my indebtedness, for many valual)le suggestions and much helpful ad\'ice, to Dr. Henrv C. Cowles and Mr. George D. Fuller, of the University of Chicago, 576 under whose joint supervision the main part of the investigation was pursued, and also to Dr. J. I\I. Greenman, of the ^Missouri Botanical Garden, for certain assistance in taxonomy. Gener.^l Fe-^tures of Skokie Marsh Skokie Marsh* is intimately associated with Skokie Stream— a small sluggish meandering stream beginning west of Waukegan, 111., and extending southeast. Years ago this stream doubtless flowed on until it at last joined the East Branch of the North Branch of the Chicago River. Today, however, its identity as a stream is lost at a point west of Glencoe, 111., where much of the water spreads itself over the marsh or enters some of the artificial drainage ditches. Figure 17, Plate XCIV, shows a more or less artificial basin at the south end of the marsh (west of Winnetka), in which water col- lects, flowing thence southward through a ditch. Southwest of Winnetka (west of Kenilworth and Wilmette), several broad drain- age ditches may be seen. These receive much of their water, in cir- cuitous ways, from Skokie Marsh and pass it on, all of it coming sooner or later into the North Branch of the Chicago River. One of these drainage ditches is shown in Figure 14, Plate XCII. In recent years drainage and cultivation have been carried on to such an extent along the margins of the marsh that its areal limit? can be defined only arbitrarily. As shown in the accompanying map (PI. LXXXVI, Fig. i) however, it is approximately 12 km. long, and at its southern end becomes 1.5 km. wide. For the naturalist, access to the marsh may be had at all times by means of the several roads running east and west directly across it. The scenery along certain of these roads (PI. XCIII, Fig. 15) is particularly pleasing. During spring and autumu, the ditches running along either side of the roads are usually filled with water. In some of these the water is deep enough to permit the passage of a small boat. In the spring of 191 2, when the marsh was in many places under water, a boat (PI. XCIII, Fig. 16) was found ver)- convenient for pene- trating to the interior. In earlv postglacial times, the marsh was an embayment ( Atwood and Goldtiiwait, '08, p. 58), which later disappeared and gave place to a svstem of drainage. At present the surface soil almost through- out the marsh consists of a black muck or partially decayed peat, I m. or less in thickness. Underneath is a subsoil of glacial clay. *For many additional data and photographs of Skokie Marsh, see Baker ('10). who has given also an account of its zoological aspects, with special reference to the moUuscan fauna. 5W General Features of the Marsh Vegetation Upon analysis, the vegetation at Skokie Marsh is found to con- sist of tliree rather pronounced formations.* Along the course taken by Skokie Stream, the plants constitute distinctly a reed swamp formation (PI. LXXXVII, Fig. 2). Extending along on either side of the reed swamp is a broad level expanse, intermediate between reed swamp and meadow. This mav be designated swamp meadow (PI. LXXXVII, Fig. 3; PI. XCIV, Fig. 18; PI. XCV, Fig. 19). At the outer edges of the swamp meadow, in narrow areas that have not been too much disturbed by cultivation, true meadow is commonly present. At certain places, however, there is an abrupt transition from swamp meadow, or even from reed swamp, to forest. Such a case is shown admirably in Plate XCV, Fig. 20, which pictures a small piece of forest containing Qiierciis rubra, Q. iiiacrocarpa, Q. alba, Juglans nigra, etc., separated from a branch of the reed swamp by a distance of only about 15 m. In the reed swamp the plants belong to five easily recognized as- sociations. Where the stream is deepest (as in PI. LXXXVII, Fig. 2), aquatic or amphibious species, such as MyriophyUiim hitmUc,\ M. hctcrophylluiii, Raiiuiicidiis dclphiiiifolius, and Fotaiiiogcton (cos- ferifoliiis?) are common near the center. In the shallower parts, the species are supplemented or replaced by Polygointiii MiiJilciibcrgii. P. hydropipcroidcs, J'cronica AuagalUs-aquatica, Radiciila aqitatica, Siiiiii ciciitacfoliuiii, Sfarganiniii enrycarpiiin, Glyccria scptcntrio- nalis, Alisiiia Plaintago-aqiiatica, Riimc.v I'crticiUatits, CaUitriche het- erophylla, and C. palustris. As Polygonum- hydropipcroidcs and Sium cicufacfoliuui are among the most abundant stream plants and ap- pear to be dominant, we may classify the plants growing in the stream or upon its bed, except along the margins, as the Sium-Polyg- oiiuin association; or, using Schouw's method of nomenclature (Schouw, '22, pp. 14S-150), we shall call this the Sio-polygo>ictiiuL On either side of the Sio-polygonetum a narrow or sometimes broad girdlet of Nyinphaca advcna and Castalia odorata occurs in many *The words "formation" and "association" arc nsed tliroughout this paper in the sense accepted by Warming ('09, pp. 140, 144), t.MI plant names given in tliis paper conform, iniless otherwise noted, witli the nomenclatnre of Gray's Manual (see Rohinson and Fcrnald, 'o8). fThe word "girdle" is here equivalent to the "zones" of many recent authors, and conforms with the recent proposal of Flahault and Schriiter ('10), except that it is here nsed for "bands" that are not "concentric." Professor Schniter kindly informs me by letter that this use of their word is perfectly justifiahlc. and further says, "we should have made provision for such a use."' 5^8 places along the stream. Usually these species are accompanied by species characteristic of the Sio-polygonetum ; but the soil and light conditions present in the girdles of Nymphaea and Castalia are pecu- liar to them and justify their treatment as a separate association, the Nyinpliacctmii. Landward from the Nymphaeetum are found dense and either intermixed or almost pure growths of Typha latifolia, Sparganiitm eiirycarpimi, Scirpus Hitviatilis, and S. validiis. Scat- tered to a varying extent among these species are Sagittaria latifolia and Siniii ciciitacfoliuin. Here and there are a few isolated patches of Dulichiiun aniiuUnacciun, of Decodon I'erticillatus, and of certain other species. This association will be referred to as the Scirpo- typhctiim. Again, in certain parts of the reed swamp, at stations slightly less hydrophytic, Phragmitcs communis is prominent. It forms exceedingly compact, nearly pure colonies that may reasonably be treated as an association, the Phragiiiifctum. Finally, we must mention the many large but somew'hat scattered patches of Iris ver- sicolor and Acorns Calamus, occurring in the outer parts of the reed swamp and often extending into the swamp meadow fonnation. These constitute an association of a very definite stam\), the Irido-acoretuni. A general comparison of the reed swamp associations shows that in the Sio-polygonetum and Nymphaeetum, where hydrophytism is greatest, the dominant plants are dicotyledonous. In fact, of the 15 species found to any considerable extent in these two associations, the 10 most abundant (Sinm cicntacfolinm, Polygonum hydropipcr- oidcs, P. Muhlcnbergii, Nymphaea advena, Castalia odorata, Rumex vcrticillatus, Veronica Anagallis-aquatica, Myriophylhim huinile* Callitriche palustris, and C. heterophylla) are dicotyledons.f In the other three associations the most abundant species are chiefly mono- cotyledons. The swamp meadow differs from the reed swamp in being more uniform, owing to greater parallelism between the water-table and the soil surface, and does not admit of logical subdivision into associations. The plants are principally such grasses as Calauiagrostis canadensis, Glyccria nerz'ata, Phalaris arundinacea, Poa triflora, Spheuopholis pallens, and Agrostis perennuns. These are frequently interspersed with Cares Inpuliforviis, C. z'esicaria monile, C. riparia, Scirpus atroz'ireus, S. Eriophorum, etc. The swamp meadow is used by farmers of the district for the production of marsh hay, and many of them customarily burn over the areas in the late autumn. ]\Iost of the shrubs and young trees are killed in this way, and so forest *But see Nos. 136 and 137 in Annotated List. tSee Henslow ('11). however, regarding the supposed monocotyledonous na- ture of Nymphaea and Castalia. 579 development is hindered. Trees occur only in small groups, con- sisting chiefly of Salts (S. fragilis, S. nigra, and other species), Fraxiiiiis nigra, P. aincricana, Populiis treintiloidcs, and Ulinus a>ncri- cana. Frequently associated with these are such shrubs as Cornus stolonifcra, Ccphalanthns occidciifalis, and Sanibticus canadensis. Throughout the reed swamp and swamp meadow are many spe- cies which, though very abundant, share only to a small extent in giving to the several associations their distinctive appearance. Thus, Ludvigia palustris, Proserpinaca palu^tris, Pcnthorum scdoides, and Stcnopliyllus capillaris are low in habit and obscured by taller plants in the sliade of which they may thrive. Again, Aster Tradescanti, Boltonia asteroides, Lobelia cardinalis, Teucrimn occidentale, and ScntellaAa galericiilata, while extremely common, are nevertheless conspicuous only during the latter part of the summer. The names of such species are here reserved, so far as possible, for the annotated list of species, at the end of this paper. The meadow formation, as already stated, is narrow and more or less interrupted. The soil surface slopes mildly upward, away from that of the swamp meadow. The vegetation is much diversi- fied at different places and from month to month during the vegeta- tive season. Poa pratensis and Agrostis alba are the dominant grasses, but Danthonia spicata and Agropyron caninion are frequent. Scat- tered among the grasses are Carex stipata, C. zntlpinoidea, C. scoparia, and Eleocharis palustris. In some parts of the meadow Viola cucid- lata, V. papUionacea, Senecio aureus, and S. Balsamitae are con- spicuous in ]\Iay and June, while later such species as Liliuni cana- densc and Rudbeckia liirta are the most noticeable. The stretches of forest present in many places at the edge of the marsh, while not usually considered as belonging to the marsh, are of interest l^ecause of the light that they throw upon the suc- cessional development of vegetation with the passing away of marsh conditions. Along the east side of the marsh, the ground surface slopes gently upward toward a rather high morainic ridge that roughly parallels the marsh ; and as one proceeds toward this ridge, he leaves behind him such woody species as Ccruus stolonifcra, C. Ainoniuin, Cephalanthus occidentalis, and Salix longifolia, and passes in turn thickets composed of Sanibucus canadensis, Populus trcniuloidcs and taller species of Salix, forest composed largely of Qucrcus bicolor, 0. rubra, Praxinus nigra, P. anicricana, and Ulnius aniericana, fmMy reaching forest composed of Qucrcus rubra and such upland species as Q. alba, Q. coccinca, and Carya ovata. 580 Certain Ecological Factors Livingston, in liis well-known studies of transpiration, found that, in a general \va}', the measure of transpiration in plants was fairly indicative of their respective environmental conditions. The tran- spiration rate for most plants being roughly proportionate to the rate of evaporation of water from a partially open receptacle, he introduced the porous-cup atmometer for measuring the evaporation rate of water. Four of these atmometers* were set out May 21, 191 1, at different stations indicated on the map: an instrument at station i, near the edge of Skokie Stream; one at station 2, in the outer part of the reed swamp; one at station 3, in the outer part of the swamp meadow ; and one at station 4, in a stretch of forest east of the marsh. Instrument No. i was in the center of a dense growth of Typlta latifolia. As the summer advanced, plants of Scutellaria galcricitlata and Tencrium occidcntalc grew up in the shelter of Typha. No. 2 was surrounded by Iris i'ersicolor,Siiiui ciciitacfoliiiiii, and a few plants of Typha. No. 3 was in a dense growth of Cala- niagrostis canadensis, and No. 4 in a small area of pastured forest, composed chiefly of Qiiercus hicolor and Fraxinus anicricana, but with a moderate proportion of F. nigra. The unglazed part of each porous-cup extended from about 22 cm. to about 28 cm. above the ground, gi\ing a mean height of 25 cm. Readings were taken weekly, up to and including October 15, 191 1. After correction according to the method outlined by Livingston, they were plotted graphically,! appearing as shown in Plate LXXXVIH, Fig. 4. The ordinates represent the number of cubic centimeters of water lost per day by a standard atmometer, while the abscissas represent the intervals be- tween the weekly readings. A study of this figure (4, PI. LXXXVIH) shows the periods of maximum and minimum evaporation to have been fairly harmonious at the four stations. And, again, the evaporation rate for the center of the reed swamp (Fig. 4, a), where hydrophytism is greatest, was usually lowest; in the swamp meadow (Fig. 4, c), it was somewhat higher; in the outer part of the reed swamp (Fig. 4, h), still higher; and in the Qiicrcus hicolor-Fraxinus anicricana or swamp white oak- white ash forest (Fig. 4, d), it was highest of all. These differences become perhaps even more evident if we compare the following aver- *None of the atmometers used were provided with a rain-excUidiiia device, such as is recommended by Livingston ('14). tA summarized account of these results first appeared in the Botanical Gametic (SherfF. '12), and later a more complete account, substantialUy as presented here, was puhlislicd in the Plant World ( Sherff, '13). 581 age daily evaporation amounts for the several stations for the entire period of 147 days: a, 3 cc. ; c, 4.2/ cc. ; b, 4.5 cc. ; and d, 7.91 cc. Or, taking the rate for d as 100%, then the rate for a was 38% ; for c, 54%; and for b, 57%.* Expressed in general terms, the evaporation rates were inversely proportionate to the hydrophytism of the station. This is due chiefly to the greater amount of moisture in the air where the station is hydrophytic; and again, the greater amount of atmospheric moisture was due, in many places, not merely to the greater sources of supply (soil moisture or surface water) but to the more difficult means of escape (because of the tall rank vegetation evoked b\- h}-drophytism). It will be noted that the aver- age rate in the outer part of the reed swamp (b) slightly exceeded that in the swamp meadow (c). This may be explained easily, how- ever, by the fact that in the swamp meadow the vegetation remained more dense and compact in late summer than in the outer part of the reed swamp, thus retarding evaporation. Transeau ('08) has obtained in a mesophytic forest on Long Island, N. Y., an average daily evaporation rate of 8.3 cm. This was based upon readings taken during a period of less than one month. J\Iore recently. Fuller ('11) has obtained for typical meso- phytic forest, based upon readings extending o\er 155 days, the average daily rate of 8.1 cc. While we are not justified by the data at hand in attempting final compai'isons, yet, so far as they go, these data indicate that evaporation is slightly less rapid in the swamp white oak - white ash forest than in climax mesophytic forest. If this indication is sustained by further study, as it undoubtedly will be, it will coincide c|uite closely with the fact that in the normal de- velopment of mesophytic forest from hydrophytic formations Qtier- ciis bicolor, fraxinus aiiicricana, F. nigra, etc., are antecedent to trees of the climax mesophytic type (Fagus grandifolia, Acer sacchantni, etc.). By way of comparison, it is interesting to note here the very recent paper of M'Nutt and Fuller ('12), in which the oak-hickory forest association is maintained (because of its intermediate evapora- tion rate) "midway between the black oak dune association and the *In interpreting these data, however, allowance mnst be made for the fact that in different associations the percentage of species which start out each year in the delicate and hence more critical seedling stage, varies. For yonng seedlings, dependent as they are upon their own photosynthetic activity for food, growth up to aUout 2.j cm. (the height at which these comparative readings were taken) is accompanied undoubtedly by much more risk than is the growth of young shoots from old, well-established perennial rhizomes, bulbs, tubers, etc. Hence the evap- oration rate for an entire association can not show with precision the extent to which each species, as such, is influenced during its most critical stages, viz., the first seasonal growth of its aerial shoots. 582 climax beech-maple forest, the position already assigned to it by Cowles and others in the forest succession of Indiana and Illinois." In the autumn of 191 1, a study of evaporation at different levels above the soil surface was made. Beginning September 3, weekly readings were taken with four atmometers arranged at different heights in a dense growth of Phragaiites 'couununis, and with three atmometers added to the one already at station I, among Typha. The last readings were taken on October 22. After correction to correspond with the readings of a standard atmometer cup, the data were plotted graphically. Among Pliragmites (PL LXXXVIII, Fig. 5) the average daily evaporation for the 7 weeks, at o cm. (the soil surface), was 2.5 cc. ; at 25 cm., 4 cc. ; at 107 cm., 5.3 cc. ; at 198 cm., in the uppennost atmospheric stratum among the Pliragmites plants, 7.5 cc, or just three times as great as at the soil surface. Among Typha (PI. LXXXVIII, Fig. 6)*, the average daily evaporation for the 7 weeks, at o cm., was .64 cc. ; at 25 cm., 1.5 cc. ; at 107 cm., 2.7 cc. ; at 175 cm., in the uppermost stratum, 6.4 cc.—or just ten times as great as at the soil surface. These differences in the rates among Typha were strongly accentuated because the readings were taken in autumn, when many of the Typha leaves had started to wither and bend over, thus giving greater exposure in the upper strata and greater shelter in the lower. Then, too, numerous plants of Scutellaria galericulata, Teiicrium occidentale, Polygonum Muhl- ciiberc/ii, etc., absent among Phragmitcs, were present among Typha and acted as a further check to evaporation in the lower strata (in which, to a very great extent, they vegetated). The data plotted in Figures 5 and 6, Plate LXXXVIII, cor- roborate very emphatically those of Yapp ('09), who found that during a total of alx)ut 15 days, the evaporation rate just above (not, as at Skokie Marsh, i)i the upper strata of) tall "sedge vegetation" was over fifteen times as great as it was at 12.5 cm. above the soil surface. They conform likewise with the more recent results of Dachnowski ('11), who obtained during about five days, at a height of 150 cm. in an American bog, an evaporation rate twice as great as at a height of 7.5 cm.; also with those of Fuller ('12) who ob- tained, during six months at a height of 2 m. in climax mesophytic forest, an evaporation rate 2.34 times as great as at a depth of 4 m. below the forest floor, in a ravine. Obviously, we must conclude, *Because of the faulty working of the atmometer at cm., the results for the first two and the last weeks are not plotted, and the average here given (.64 cc.) is for the remaining four weeks. Enough certain data were olitained however for the other three weeks to show that the total average would have been even less than .G4 cc. 583 witli Yapp, that plants may grow in proximity to each other and yet, if vegetating in different strata above the soil surface, be subject to widely different growth conditions. Thus, for example, Riccia na- fans and Typha latifolia, which may be found together in great quan- tity but vegetate mostly in dift'erent atmospheric strata, live under evajwration conditions differing much more than do those under which TcHcriiim occidciitalc (of the reed swamp) and Aster salici- foliits (of the swamp -white oak - white ash forest), plants of similar height and growth form, live. The depth of the water-table in the reed swamp and the swamp meadow was observed each week from May 21 to October 22, 191 1. The water in Skokie Stream was about i m. deep in ?^Iay, after which it gradually declined until in July, when the stream bed was in most places fairly dry. In August the water began to rise again, and by October had reached an average depth of about i.i m. In the rest of the reed swamp and in the swamp meadow the water- table during May was coincident with or above the soil surface ; thereafter it sank, until in early September the maximum depth of I m. in the reed swamp and 1.75 m. in the swamp meadow was reached ; and then, rising rapidly, it reached the surface again by the middle of October. During 191 2, water was much more abundant throughout the marsh. Seldom could the reed-swamp be traversed without the use of boots, even in midsummer. According to farmers in the vicinity of Glencoe, Skokie Stream has sometimes in the past risen until a depth of about 3 m. was reached, when the entire marsh was of course deeply submerged. Various attempts have been made to classify the constituent species of a formation with relation to the optimum water-table depth for each species. But where the water- table varies greatly in depth from month to month and from year to year, data must be secured through many years if they are to show more than merely the relative degrees of hydroph\i:ism to which plants in different places are subject. Litmus tests each week, from May 21 to October 22, 191 1, showed the water in Skokie Stream to be either neutral or slightly alkaline. Similar tests showed the soil water in the outer parts of the reed swamp and in the swamp meadow to be usually neutral or slightly al- kaline, except that for a few days in August acid was present, al- though the amount was almost negligible. SUBTERRANE.\N OrG.ANS AND THEIR InTERREI^ATIOXSIIIPS A study of the subterranean organs of the reed swamp plants showed that in many cases their depth is roughly proportionate to 584 the depth of the water-table. Yapp ('08) arrived at a similar con- clusion concerning the plants at Wicken Fen. And since the depth of the water-table may influence the depth of the subterranean or- gans, the latter in turn may be a potent factor in the success or failure of various species. Thus, for example, the rhizomes of Polyg- oniiiii MiihJcnbcrgii, where this species occurs in the Sio-polyg- onetuni are usuall}- at or near the surface of the stream bed. As King ('97, p. 240) and others have pointed «>ut, saturated soil like that of the stream bed does not admit oxygen freely ; and so in the Sio-polygonetum, the rhizomes of Polygomtui and their roots appear advantageously placed. But in the Scirpo-typhetum (PI. LXXXIX, Fig. 7), where the surface soil is occupied by an extremely dense mat composed of the rhizomes of Typhd, Sparqanhtjii, and Scirpus, the rhizomes of Polygonum average about 10 cm. in depth; hence in the Scirpo-typhetum, although the rhizomes of Polygonutn are lower, evidently in response to the greater average depth of the water-table, they have the additional advantage of being able to travel with less interference from the other rhizome systems. An examination of Typha, Sparganiiim, Scirpus Hu-ciatilis, and S. zvlidiis shows these species to be very similar in growth-form and hence capable of keen competition. Where any one of these species becomes more abundant in the Scirpo-typhetum, the others become less so. Because of the thick, strong rhizomes, the subterranean competition is to some extent mechanical ; but it is probably to a much greater extent, as Clements ('05, pp. 285-289) maintains, physi- ological (or "physical"), especially in the case of the roots proper. The opposition that any or all of these species can offer to the intru- sion of other species makes their hold upon the soil very effective. With Sagittaria (PI. LXXXIX, Fig. 7), however, the case is dif- ferent. Its growth-form favors a less compact arrangement of the individual plants, as its rhizomes can not produce a thick mat. Ob- viously, as the plants of Sagittaria are developing vegetatively, other species, such as Typha, Sparganiuiu, and Scirpus, may easily invade and occupy the soil with their densely matting rhizomes. Subse- quently the rhizomes of Sagittaria, if they are to establish new plants at proper distances away from the parent plant, must either plough their way through the surface mat of rhizomes or travel underneath it. They usually do the latter. As a rule, several rhizomes start growth from each plant in early summer in a downward direction ; at a depth of 10-15 cm. they assume a horizontal direction for some distance ; and then grow upward again, with a tuberous, propagative thickening near the distal end, and finally resemble somewhat a shal- 585 low, inverted arch.* Thus, interference fruni surface rhizomes and roots is to a great extent avoided. In this case, then, while it is not certain that the inverted arch of the Sagittaria rhizome is a direct adaptation to this particular struggle, it is certain that it is here of the greatest value, however induced originally. Pieters ("oi) found among the plants of western Lake Erie that even where Sagittaria latifolia was most abundant, Spargaiiiiiin (and Zisaiiia) had secured a foothold. On the other hand, throughout all the broad "zones" of Sparganiiun, Scirpits validiis ("S. lacitstris"), and 5". flin'iatilis that he describes, he says Sagittaria latifolia wa.s common. Thus, in these cases, Sagittaria was found able to asso- ciate successfully with Sparganiiun and other species having a Spar- gaiiiiim growth-form, even where these species formed dense "zones". A study of the subterranean organs of Sagittaria, Sparganinm (or Typlia or Scirpns), and Polygonum shows that because of differences in direction or in depth they conflict but little. Again, because of ditferences in growth-form, their aerial parts do not conflict seriously. Thus a given area can usually support a greater mass of vegetation if these three growth-forms I)e present in fair mixture than if only one be present. Spalding ("09) has described the mutual relation- ships of Ccrctis gigantcns and Parkinsonia microphylla, two desert species which thrive together because the occupation of different depths by their root systems enables them "to utilize to the utmost the scanty rainfall." Woodhead ("06) found Holcus, Pteris, and Scilla forming a noncombative "society or sub-association." For a group of plants mutually competitive, Woodhead uses the term "com- petitive association." Recently Wilson ('11) likewise speaks of a "com]:)lementary association" or "society." But the use of the words "association" and "society" in this connection is unfortunate. These words have been used already by Cowles ('01) and others (see Warming '09, p. 144) to denote a primary subdivision of a forma- tion. As will be seen later (and in fact as Woodhead's interchange- able use of "sul>association" and "association" might imply), not all complementary or competitive groups are coextensive with a true association. We shall here substitute the word community, which is of less restricted application. Thus Sagittaria and Polygonum, where occurring in the Scirpo-typhetum with either Typha or Scirpns fluviatilis or S. validiis, constitute a complementary community; but Sparganinm, Typha, Scirpus HuziatiHs, and .S". x-aliilns, where they occur intermixed, form a competitive comnumity. *For illustrations of the .similar rhizomes of Sagittaria sagittifolia see Gliick ('05, pi. C and figs. 35, 39). 586 Species that are plainly complementary in one association may be less so in another. Thus, Polygomiin Muhlcnbcrgii and Sparganium are complementary in the Scirpo-typhetnm ; but in the Sio-polygone- tum, where their rhizomes lie in common near or at the surface of the stream bed, they are "edaphically" (see Woodhead, "06) com- petitive, and hence complementary only in an aerial way. In this particular case, however, the frequently open appearance of the vege- tation in the Sio-polygonetum indicates that the mutual biotic strug- gle of the two species is less keen than their separate struggles against somewhat adverse environmental conditions. In the reed swamp certain mints become conspicuous during mid- summer, particularly so in the Sciipo-typhetum, where they thrive in the shelter of Typha and other tall plants. Tcucriitin occidcntale and Scutellaria galericulata are very common. They produce from their basal nodes numerous slender stolons that run out at different depths in the soil, and these stolons may produce new plants. These species tend to have their root systems 3-6 cm. lower in wet situations than in dry, although exceptions to this rule are not rare. But whether growing from plants in dry or from those in wet situations, the new stolons exhibit a remarkable power of changing their direction of growth, in response to numerous obstructions, and thus they may proceed further without serious results. Considering the strength and size of the rhizomes of Typha, Spargaiiiuin, and Scirpus, also the delicate nature of the stolons of Tciicriiiuh and Scutellaria and their capacity for altering growth-direction, it is probable that me- chanical competition between such rhizomes as those of Typha and such stolons as those of Teucrium is practically absent. Again, the aerial parts of the Typha form vegetate chiefly in higher atmospheric strata than do those of the Teucrium form. Evaporation readings show that in higher strata evaporation is much greater; and while plants of relatively xerophytic structure (e. g., Typha, Sparganium, and Scirpus) are fitted to withstand acute drying conditions, plants with foliage of looser texture (e. g. Teucrium and Scutellaria) can vegetate better in lower strata, where the effect is that of greater humidity, the abundance of the latter plants among the former at Skokie Marsh tending to confirm this statement. Further, the per- sistence with which tall plants like Typha become dominant under favorable soil conditions shows that they are not, at least not notice- ably, harmed by plants like Teucrium. If, finally, we allow for the great availability of nitrogenous foods in the soil and for the dif- ferences in food rerjuirements, it becomes clear that the numerous 587 communities of Typlia and Tcnci^iiiiii, Typha and Scutellaria, Spar- ganiiiiii and Tciicriiiiii, etc., are complementar)-. The purity of the Phragmitetum has already been mentioned. Many species that flourish elsewhere in the reed swajnp under a wide range of light, moisture, and other shelter conditions fail to thrive here. Only Calaiiiagrostis ca)iadensis gains noticeable en- trance, and then imperfectly. The dead Phraginitcs, the growth of previous years, makes a considerable but loose covering near the soil, its decay not being facilitated as in the Scirpo-typhetuni, where water is more abundant. This dead cover may perhaps act as a partial check upon the invasion of other species; but a study of the rhizomes of Phraginitcs (PL LXXXIX, Fig. 8) shows another fact which probably is more important. They do not occupy one particular level, but rather several dift'erent levels of soil. As a result, there is formed a dense mat of rhizomes and roots, about 2.5 dm. deep. Ob- ^iously, the subterranean organs of other species which might start growth here must compete with the extraordinarily large number of Phraginitcs roots and rhizomes. Where other factors are suited ecpially to Phraginitcs and to competing species, this biotic factor in the subaerial struggle ought usually to be decisive in favor of Phrag- initcs. No cases were found where Phraginitcs had regularly produced rhizomes (or stolons) upon the surface of the ground. Frequent in- stances were met with, however, in which the entire aerial shoot had fallen o\er upon wet, mostly nude soil and, having produced numerous roots, had elongated at a much more rapid rate than be- fore. The N^ymphaeetum displays many complementary communities. The rhizomes of Nyinphaea advcna (PI. XC, Fig. 9) are usually 5-10 cm. thick and lie mostly at a depth of 8-25 cm. below the soil sur- face. The rhizomes of Castalia odorata, while smaller, lie at a simi- lar depth. Where the Nymphaeetum intergrades with the Scirpo- typhetmn, as is commonly the case, the rhizomes of Typha, Sparga- niiiin, and Scirpiis ixilidiis lie higher in the soil. In many places the soil surface itself is occupied by the stolons of Ranunculus delphini- folius and the creeping stems of Polygonum hydro.pipcroidcs, with a large, upright stem base of Siuin cicutacfolium present here and there. In other places. Ranunculus is replaced by Myriophyllum humilc or by young plants (growing chiefly from detached leaves) of Radicula aquatica,\\\\\\Q Polygonum is replaced by J'cronica Ana- gaUis-aquatica, and Stum by Rumex vcrticillatus. And while it is true that Nyinphaea and Castalia, or Typha and Sparganiuin and 588 Scirpiis, or Ramiiiciilus and Myriopliylliiin and Radiciila, or Polyg- oiniiii and Veronica, or 5i!(»J and Riimc.v are mutually competitive, yet a complete community (as shown, e. g., in PI. XC, Fig. 9) is complementary ; the basal parts chiefly because of different depths, and the upper parts chiefly because of different growth-forms. An inspection of the Xymphaeetum shows that only where Nym- pliaca is nearly or quite absent does Sagittaria latifolia successfully in\'ade from the Scirpo-typhetum. As is commonly known, the rhi- zomes of Ah'iuphaca in many habitats are usually decayed to within a short distance of the growing apex. An investigation during Au- gust, 191 1, showed that generally where the rhizomes of Sagittaria had penetrated these decayed parts, they themselves had started to decay.* Frequent cases were found where the decayed Nyinphaca rhizomes lay nearer the surface and the Sagittaria rhizomes had pro- ceeded underneath, unharmed. In many instances, however, where the stem-tubers had been mechanically impeded (by woody roots, etc.) in the encasing soil, they had decayed. And here, while the decay must have been due to some one or more physiological causes, yet these causes could not have operated had not mechanical impedi- ments first retarded the stem-tubers for a sufficient length of time. As our knowledge of the interrelationships of subterraneaia organs progresses in the future, we shall prol>ably find that often, in the case of certain species with large subterranean parts, there is offered or received mechanical resistance which is immediately decisive in com- petition because of the physiological processes that it promotes. Speaking in a general way, while Nymphaca and Sagittaria thrive better in the Nymphaeetum and Scirpo-typhetum, respectively, yet along the line of ten.sion between these two associations the injury done by the decayed Ah'iiif^haca rhizomes to the rhizomes of Sagit- taria is a factor that appears to be decisively in favor of Nyinphaca. The inverted rhizome arch of Sagittaria, useful in the Scirpo-typhe- tum, is here more often harmful. In many parts of the Irido-acoretum, Pol\gonuiu Mulilciibcrgii and Galium Claytoui abound, and tliese form with Acorns a comple- mentary community (PI. XC, Fig. 10). The creeping stems of Galium, root upon the soil surface, the rhizomes of Acorns lie just beneath, and those of Polygonum are deepest of all. The bushy "hoot of Galium appears not to hami the slender, ensiform leaves of Acorns, and they in turn do little harm to it. In late summer, the *Many litmus tests uniformly showed the decayed parts of the Nytiiphaca rhi- zomes to be strongly acid. Enough cultural experiments have not been performed, however, to determine whether the effect upon the Sagittaria rhizomes, as above noted, was due to acid or to other causes. i 589 shoots of Pulyyoniiiii rise above those of Aconis and Galium witli- out apparent harm to either of them. And while Polygoitiiin might increase in abundance if Acorns and Galin in were entirely absent, still to a great extent the community, viewed as a whole, is comple- mentary. Elsewhere in the Irido-acoretum the rhizomes of Acorns are replaced by those of Iris; and very often the rhizomes of Galiniii are replaced by those of Lndz'igia palnstris, L. polycarpa, Proser- pinaca pahistris, Pcnthoruin seJoidcs, Veronica scntcllata. or Cain- paivda aparinoidcs. The basal parts of the various swamp ,meadow species are usually more slender than those of the reed swamp species, and hence the texture of the surface mat of rhizomes, roots, etc., is finer. Then, too, reproduction by seeds becomes more common. Polygonnm Miililenbcrgii is present in the swamp meadow, and by means of its extensiveh' creeping rhizomes, which lie rather low, it forms in some places large patches. Certain other perennials, e. g., Asclepias in- carnata and Sinin cicittacfoliuui, which root near the surface, may reproduce largely by seed or by new shoots arising from the old stem base of the preceding year. In the middle and latter parts of the summer, when the surface soil is no longer saturated with water, such annuals as Panicuiu capillarc, BcJiinocldoa crusgalli, Erar/rostis hypnoidcs, Stcnophyllns capillaris, Polygouiiin Pcrsicaria, Acnida sp. (see Annotated List, No. Sg), Amaranthns panicidatus, and Brechtitcs Iticracifolia take possession of all exposed surface soil and become ex- ceedingly abundant. Much of the surface soil that has been denuded by burning or by other causes is already occupied, however, by the rhizomes of ]ierennials such as Lndvigia pahistris, L. polycarpa, Proscrpiiiaca pahistris, etc. In these cases Boltoiiia astcroidcs, Cal- litrichc hctcrophyUa, and C. pahistris are often abundant. Both spe- cies of Callitriclic, however, die away in midsummer, being replaced by annuals. Figure ii, Plate XCI, shows such a community. Calli- trichc, maturing earliest, is "seasonally" fWoodhead '06) comple- mentary wMth the other species. BoUonia roots lowest, while its aerial shoot grows much the highest ; and since it is not harmed very much by Proscrpiiiaca, Ludvigia, and Pcnthornm, while they derive, if am 'v"^ ig, benefit from its shelter, Boltnnia is complementary both aerialj '^ '^ \ subaerially. Proscrpiiiaca, Budvigia, and Pcntho- ruin are vt.', similar throughout in growth form and they consti- tute mutually a competitive community; but, even though mutually competitive, they form with Boltoiiia and Callitriclic a community that may properly be called complementary. 590 As has been stated already, the flora of the meadow is highly di- versified. A very large number of definite interrelationships, similar to those detailed for the reed swamp and the swamp meadow, are found to exist, but lack of space precludes more than a brief descrip- tion of a few examples. In the moist parts of the meadow, the soil at a depth of 3-1^ cm. frequently contains the tuberous thickened roots of Cicuta maculata and Oxypolis rigidior, and also the tuber- bearing rhizomes of Bqnisctnin arvense. In drier situations the bulbs of Liliuiii canadensc occur at a similar depth (most often about 10 cm. deep). Higher in the soil may be found (PI. XCI, Fig. 12) roots of such species as Asdcpias iiicarnata, Thalictnim rcvoluhiin, and Lathyrus palustris, while the surface soil contains a mixture of the root systems of Poa pratcnsis, Agrostis alba, Blcocharis palustris, Acalvpha I'irginica, etc. In the community shown in the figure just mentioned, Equisetum is edaphically complementary, but (considering only the aerial sterile shoots) aerially competitive with Poa, Agrostis, Blcocharis, and Acalypha. To a moderate extent, the plants rooting near or at the surface appear to be complementar>' with the plants rooting deeper. Small, apparently open depressions are numerous in the moist parts of the meadow. These generally contain (PI. XCII, Fig. 13) such plants as Iris, Aconts, J 'iola conspcrsa, V. cucullata, V. papili- onacea, Cardaniinc bidbosa, and seedlings of Lycopus aiiiericantis. And while the rhizomes of Cardaniinc and Lycopus occur almost in- variably just below those of the other species, and while the different species doubtless mal-ce different demands upon the soil, yet edaphic competition is undoubtedly sharp. Their rhizomes are mostly short and thick, lie just below or at the soil surface, and form a dense mat. Nevertheless, when one or more square feet of this mat were care- fully removed and the soil in the interstices among the rhizomes was taken away, it was estimated that the interstices, as viewed from above, constituted from 35 to 60 per cent, of the total. Evidently, then, so far as mere room was concerned, several other species could have grown—in fact, did grow—in these interstices. But they were plants which rooted higher or lower; or, if at the same level, they were species not largely dependent upon rhizomes or stolons for mul- tiplication. Thus, where Iris versicolor had reached a, maximum of frequencv, F?lygonum Muhlcni v-gii, with a low i-nsii system, and Galium Claytoni, with a high root system, might liv^to but Acorus Calamus, with rhizomes similar to those of Iris versicolor and lying at a similar depth, and dependent largely on rhizomes for multipli- cation, was absent. 591 The almost complete absence, in these small areas, of stolonifer- oiis or loosely spreading species makes it seem certain that there exists some mechanical competition in which species of compact and fre- quently cespitose habit or species capable of reproducing extensively from seed are successful. The extent, however, to which their suc- cess is achieved because of their growth-form or because of their superior adaptation to the particular complex of soil and moisture conditions in these small areas, is of course incapable of accurate estimation without further study. The idea of mechanical competi- tion (/. ^., a struggle either among the various species because of the mutual bodily resistance of any or all of their growing parts, or of indi\idual species because of the resistance offered by the soil's com- pactness to the locomotion of their subterranean organs) is opposed by Clements ('05, pp. 285-289) ; but Warming ('09, p. 324), in ac- counting for the usual absence of vegetative locomotion among per- ennial herbs of the meadow formation, seems inclined to accept this idea in part. Summary and Conclusions 1. Atmometer readings at a uniform height of 25 cm., taken for a period of 147 days at four different stations, show that the evapora- tion rate is lowest in the center of the reed swamp and gradually in- creases as conditions approximating those of forest are reached. 2. The evaporation rate found to obtain in the swamp white oak - white ash forest, conforms with the commonly known fact that with successive increases in the mesophytism (attended with decreasing hydrophytism) of a forest, trees such as Qiicrcus bicolor, Fra.vinus nigra, and F. anicricana are antecedent to trees like Fagits grandi- folia and Acer sacchariim. 3. Atmometer readings, taken for seven weeks at four different le\els among Phragmites plants and at five different levels among Typlia plants, show that among marsh species of compact social growth evaporation is proportionate to the height above the soil. These results thus coincide with those of Yapp ('09). 4. Data accumulated at Skokie Marsh support the conclusion of Massart ('03) that it is a matter of importance to perennial plants that their hiljernating organs occupy a definite level in the soil. 5. Certain observed cases of variation in this level (Teiicrinm occidentale, Polygomtm Mithlenhergii, etc.), corresponding to changes in the water-level, indicate that with certain species, at least, the depth of the water-table is much the most potent controlling factor (cf. Yapp, '08). 592 6. Two or more species may live together in harmony because (i) their subterranean stems may lie at different depths; (2) their roots may thus be produced at different depths; (3) even where roots are produced at the same depth, they may make unlike demands upon the soil; (4) the aerial shoots may have unlike growth-forms; or (5) even where these growth-forms are similar, they may vegetate chiefly at different times of the year. According as one or more of these conditions control the floristic composition of a given com- munity the community may be called complementary. 7. The root depth having been determined by various factors for the different species in a community, the specifically different root systems then fvmction in a complementary or a competitive manner as the case may be. But even if the root systems be complementary, the community may be competitive because of marked competition among the aerial parts. Likewise, competitive root systems may render competitive a community otherwise complementary. 8. Through the ability of certain species to utilize different strata in the soil, the aerial portions of these plants are brought into a closer competition. And with closer competition, the chances in the past for further adaptation of similar aerial shoots to dissimilar growth conditions must have been greatly increased. Hence com- munities formerly complementary in a purely edaphic way, may have been largely instrumental in the evolution of completely complemen- tary communities. In so far as they have been thus instrumental, the fact deserves great emphasis, especially when we consider the far- reaching changes in form and anatomical structure necessarily de- veloped as a prerequisite to living in a completely complpmentaiy community. Annotated List of Plant Species As a matter of taxonomic interest to botanists in the future, it seems worth while to present here an annotated list of all the species of the Pteridophyta and Spcnnatophyta foimd growing to any extent in Skokie Marsh. Stray species (especially weeds), occasionally ob- served, have not been included in the list, except where evidence in- dicated that they were regular inhabitants. Also, many weeds which occur along the roads traversing the marsh and which do not prop- erly belong to the marsh flora, are omitted. As the following list stands, then, it includes only the established species found in the reed swamp, swamp meadow, and meadow of Skokie Marsh proper. 593 Polypodiaceae 1. Aspidiuvi Thelypteris (L.) S\v. Mostly in the swamp meadow and outer parts of the reed swamp; fairly frequent. 2. Onoclea scnsibilis L. In the swamp meadow and reed swamp; rather rare. Bqiiisctaccae 3. Eqnisetmii arjense L. . , - In the meadow ; frequent. / Typhaceae 4. Typha latifolia L. In the reed swamp ; abundant. Sparganiaceae 5. Sparganium ciirycarpiim Engelm. In the reed swamp; very abundant, in many places almost choking up the streams and ditches. In late summer, 191 1, after the water in Skokie Stream had fairly well disappeared, great quantities of young aerial shoots were put forth by this species where it occurred upon the stream bed. When these shoots flow- ered and fruited, they manifested a striking appearance, in that they were yellowish green in color,—not dark green, as was the species elsewhere (cf. Harshberger, '04, p. 136). Najadaceae 6. Potamogeton sp. Potamogeton material was fomid in 191 1 in several of the deeper places in Skokie Stream from the county line (Brae- side) road as far north as to Lake Forest. This was not in fruit, but appeared on careful comparison with herbarium specimens to be P. zosterifolius. Alismaccae 7. Sagittaria latifolia Willd. , Very common in the reed swamp and frequent in the swamp meadow. Occasionally, detached stem-tubers from this species were found being carried along slowly in Skokie Stream, indi- cating that Sagittaria may at times migrate considerable dis- * tances in a purely vegetative way. 594 8. Alisiiia riantago-aqiiatica L- In reed swamp ; common. Confined mostly to the very wet places. Graniineac 9. Panicitui capiUarc L- Along ditches, stream-banks, etc. Very common in late sum- mer. 10. Echinochloa cnisgalli (L.) Beauv. Prominent in late summer in open or mown areas. 11. Lccrsia orysoides (L.) Sw. In reed swamp and swamp meadow ; frequent. 12. Phalaris anindinacea L. In the outer reed swamp, also in swamp meadow ; abundant. Yapp, ('08, p. 67) has pointed out that the leaves of Phrag- inites, because of the slippery inner surface of their sheath, can easily turn about so as to stream with the wind. The same was found at Skokie Marsh to be true, though in a lesser degree, of Phalaris anindinacea. 13. Phlcum praicusc L. In the meadow, where it was more or less frequent, probably because of its occasional cultivation in certain fields near the marsh. 14. Alopccurus gcniciilatiis L. In the swamp meadow and outer parts of the reed swamp, west of Glencoe ; sparsely scattered. 15. Agrostis alba I,. Common in many parts of the meadow ; occasional in the more open parts of the swamp meadow. 16. Agrostis perennans (Walt.) Tuckerm. In meadow and swamp meadow ; abundant, especially west of Glencoe. 17. Calaiiiagrostis canadensis (Michx.) Beauv. The dominant grass of the swamp meadow and frequent in the reed swamp ; forms the bulk of the hay obtained in Skokie Marsh. 18. Sphenopholis pollens (Spreng.) Scribn. In the swamp meadow ; scattered in very small patches. « 595 19- Danthonia spicata (L.) Beauv. In dry parts of the meadow, west of Glencoe ; found spar- ingly. 20. Spartina Michauxiana Hitchc. Along ditches, in moist depressions, etc. ; found only rarely. 21. Phrar/iiiitcs communis Trin. In conspicuous dense patches in the reed swamp ; abundant west and southwest of Glencoe. 22. Eragrostis hypiioidcs (Lam.) BSP. Very common in the wet open places of the reed swamp ; fairly frequent in the swamp meadow. 23. B. Frankii (Fisch., Mey. & Lall.) Steud. In moist open places of the reed swamp; found sparingly. 24. Poa coinprcssa L. In the meadow ; rare. 25. P. trinora Gilib. Very common in the swamp meadow. West of Glencoe this grass forms a fair percentage of the marsh hay obtained each summer. 26. P. pratcnsis L. In the meadow ; abundant. 27. Glyccria nerz'ata (Willd.) Trin. In the swamp meadow ; common, especially west of Glencoe. Often occurring in almost a pure growth. 28. Glyccria scptcntrionalis Hitchc. Along Skokie Stream or upon the stream bed ; frequent. 29. Agropyron caniiutm (L.) Beauv. Found in only one part of the meadow, west of Glencoe. 30. Honiciim jubatiim L. Commonly with Poa triflora and Agrostis pcvcnnans, in the swamp meadow ; frequent along the roadsides. 31. Elyiiius virgiiiiciis L. Along ditches ; occasional. Cyperaccac 32. Dulicliiuiii arundinacciim (L.) Britton Only one patch found, this in the reed swamp west of Brae- side. 596 33- Eleocharis obtusa (Willd.) Schultes In the swamp meadow and reed swamp ; scattered mostly in open moist places. 34. £. palitstris (L.) R. & S. In grassy places of the meadow and swamp meadow ; com- mon. West of Winnetka this species was found growing in Skokie Stream, where it was much stouter and attained an aver- age height of .6 to .7 m.—a fact conforming wtih the observa- tions of others (cf. Gray's Manual, Robinson and Fernald, '08, p. 183). 35. Stciiophyllus capillaris (L.) Britton Everywhere in the reed swamp and swamp meadow ; com- mon. 36. Scirpus validus Vahl. In very wet parts of the reed swamp ; common. 37. S. fluz'iatilis (Torn) Gray In the center of the reed swamp, where it frequently fringes Skokie Stream in large patches. 38. 5". atroz'irciis Muhl. Scattered here and there, often abundantly, in outer parts of the reed swamp and swamp meadow. 39. S. lincatus Michx. Found with the last species, but only rarely. 40. S. cypcriiiiis (L.) Kunth. In the swamp meadow ; rare. 41. S. Eriophonini Michx. In the outer reed swamp and in the swamp meadow ; common. 42. Carcx scoparia Schkuhr. In the swamp meadow and moister parts of the meadow; common. 43. C. cristafa Schwein. Same range; common. 44. C. vulpinoidea Michx. Same range; common. 45. C. stipata Muhl. Same range ; common. 597 46. C. cnisconi Shuttlw. ' Occurring in two small patches west of Glencoe, in wet soil. Mr. E. J. Hill informs me that many years ago he found a con- siderable quantity of this species there. 47. C. aurca Nutt. In swamp meadow, west of Glencoe; found sparingly. 48. C. laniigiuosz. Michx. In swamp meadow ; in some places, covering considerable areas. 49. C. riparia W. Curtis Mostly in the swamp meadow ; abundant. 50. C. lupulifoniiis Sartwell Observed west of Braeside, in the outer part of the reed swamp. 51. C lupiilina Muhl. In the swamp meadow ; seemingly rare. 52. C. vcsicaria L., var. monile Tuckerm. Mostly in the swamp meadow ; fairly common. Note.—Several other species of Carex. among them probably C. granularis, were found, but because of their inunature condition or complete lack of flowers and fruit, positive determination could not be made. • Araccae 53. Acorns Calamus L. ' In the reed swamp; common. Le>}tnaccae 54. Spirodela polyrhiza (L.) Schleid. Abundant in some places upon the surface of the water in Skokie Stream. 55. Lcmna trisulca L. In Skokie Stream west of Glencoe; found in 191 1 and 1912, at only one place. Juiicaccae 56. /uncus tenuis Willd. In the meadow ; occasional. 57. /. Dudlcyi Wiegand Same range ; occasional. 598 58. Liliiim canadcnsc L. Lihaccac In the meadow; common, especially west of Glencoe. 59. Iris versicolor L. Iridaccac In the reed swamp and swamp meadow ; common. Orchidaccac 60. Habenaria Iciicophaca (Nutt.) Gray Observed in the meadow, west of Glencoe ; only one small colony of plants found. 61. Salix nigra Marsh. Salicaccae Here and there in wet soil, mostly along the roads. 62. S. amygdaloidcs Anders. The most abundant of the willow trees in Skokie Marsh; preferring the wet places. 63. S. alba L., var. zitelliiia (L.) Koch Occasional as a large tree in rows along ditches west of Glen- coe and Winnetka, where it was evidently planted by man. 64. S. longifolia Muhl. Frequently covering low wet depressions in the swamp meadow. Flowering in frequent cases until late autumn. (PI. XCVI, Fig. 21.) 65. S. cordata Muhl. In wet places; frequent. 66. S. discolor Muhl. In wet places ; common. S. discolor var. prinoidcs (Pursh) Anders. Several shrubs in the marsh, appearing as hybrids between S. discolor and 5". cordata, are referred to this variety. 67. S. pctiolaris J. E. Smith In meadow and outer part of swamp meadow; occasional. 68. 5*. humihs Marsh. In the meadow ; found only sparingly. 69. S. rostrata Richards West of Glencoe and Pavinia: rare. Note.—Two or three other forms of Sall.v were found, which were not typ- ical of any known species, hut appeared to he hyhrids between certain species enumerated above. In the absence of expert opinion coriceming their status, separate treatment is here omitted. Salix fragilis. occurring in the swamp meadow alone; certain roads and apparently adventive recently, is likewise omitted in the list. 599 /O. Populiis trcuuiloidcs Michx. In thickets, small patches of forest, etc. ; frequent. /I. P. grandidentata Michx. \\\\\\ the last, but rather rare. /2. P. dcltoides Marsh. Along ditches and roads; merely a few scattered trees. Urticaceae •JT,. Uliiiiis aincricaiia L. A few large trees here and there. (PI. XCVI, Fig. 22.) Polygoiiaccae 74. Rimic.v britannica L. In the reed swamp ; frequent. 75. R. crispits L. Occurring sparingh' in the meadow (but common along the roadsides). 76. R. altissiiiiiis Wood In the swamp meadow ; occasional. /J. R. vcrticillatits L. Very abundant at many po'nts in Skokie Stream ; frequent in other parts of the reed swamp. 78. Polygonum avicularc L. In the meadow, where principally along paths ; occasional. 79. P. lapathifoliiini L. In the swamp meadow ; rather common. 80. P. M'.ihlenhcrgvi (Meisn.) Wats. Abundant in the reed swamp; frecpent in the swamp meadow, 81. F. pcnnsylvanicum L. In a few open, fairly dry areas of the reed swamp; not com- mon. 82. P. Hydropipcr L. Mostly in the Irido-acoretum of the reed swamp; common, becoming very abundant at certain points. 83. P. acre HBK. Occurring with P. Hydropipcr, Init less abundant. 84. P. Persicaria L. Scattered, in open parts of the swamp meadow; fref|uent. 600 85. P- hydropipero'ides Michx. Confined mainly to Skokie Stream, in which, at some points (especially west of Braeside and Glencoe), it occurs in great abundance, to the almost complete exclusion of other species. 86. P. sagiffatian h. In the outer parts of the swamp meadow ; rare. 87. P. scandals L. Rare in the marsh proper ; confined mostly to roadside thick- ets. Chenopodiaccae 88. Cliciwpodium album L. In open spots of the meadows ; frequent. Amaranihaccae 89. Acnida sp. Very common along Skokie Stream and in more open spots of the swamp meadow. Material was originally determined ac- cording to the older manuals as A. tamariscina (Nutt.) Wood. Absence of pistillate plants among my specimens makes it im- possible now to apply positively the more precise nomenclature of Gray's New Manual (see Robinson and Pernald, '08, p. 373) ; but the Skokie Marsh plants probably belong to A. fuberculata Moq. and its variety suhmtda Wats. 90. Amaranthus paniculatus L. In open, fairly dry spots of the swamp meadow. In 1912, this species was found only rarely, and it is probable that much of the material considered in 191 1 as A. panicttlatiis was the up- right form of Acnida sp. CaryopJiyllaccae 91. Arenaria lateriflora L. About thickets in the meadow and outer swamp meadow ; somewhat frequent. 92. Stcllaria longifolia Muhl. Among the grasses and sedges of the swamp meadow; fre- quent. 93. Ccrasliinn nutans Raf. In a few open, moist, shady spots in the swamp meadow; rather rare. 601 94- Nymphaca adz'cna Ait. In the Nymphaeetuni of the reed swamp ; abundant and con- spicuous. 95. Castalia odorata (Ait.) Woodville &' Wood With Nymphaea advena, to which it appears ecologically equivalent ; common or even abundant. Ranunculaceae 96. Ranunculus delphinifolins Torr. Abundant in Skokie Stream ; the seedlings frequent in open or sheltered moist depressions of the swamp meadow. 97. R. sceleraius L. In meadow and swamp meadow ; rare. 98. R. Pcniisylz'aiiicus L. f. Here and there in the swamp meadow ; somewhat rare. 99. Thalictntni rcvoluUim DC. In the meadow ; frequent. 100. Caltlia palustris L. In the swamp meadow ; rare. CvHciferae loi. Radicula palustris (L-) Moench In the swamp meadow ; abundant. Radicula palustris, var. hispida (Desv.) Robinson Growing with the species proper, and abundant. Of the hundreds of specimens examined, none was found showing any intergradation with the species itself. 102. R. aquatica (Eat.) Robinson Abundant in Skokie Stream, especially west of Glencoe. Ap- pearing to renew itself chiefly by its detaching leaves, which take root and propagate new plants,—a habit already noted by other observ'ers. 103. Cardamine bulbosa (Schreb.) BSP. In the reed swamp, swainp meadow, and moist parts of the meadow ; common. 104. C. pennsylvanica Muhl. In the reed swamp and swamp meadow ; common. Numer- ous seedlings develop in late summer and flower until late autumn. 602 Crassiilaccae 105. Pcnthorum sedoidcs L.' In the reed swamp and swamp meadow ; common. Saxifragaccae 106. Ribcs Horidum L'Her. In thickets ; somewhat frequent. 107. R. nigrum L,. In thickets ; apparently rare. Rosaceae 108. Fragaria virginiana Duchesne In the outer meadow ; fairly frec^uent. 109. Potentilla monspelicnsis L. In the meadow and open spots of the swamp meadow. no. P. palustris (L.) Scop. Growing with Typha latifolia, west of Braeside and Glencoe; only two small patches observed. Aerialh', this species is strongly complementary with Typha; it utilizes the lower atmospheric strata, where it flourishes among the aerial shoots of Typha. 111. P. canadensis L. In drier parts of the meadow ; occasional. 112. Geinn z'irginianum L. In open places of the swamp meadow ; rare. 113. Rosa blanda Ait. In the meadow ; frequent. Doubtless one or two other species of Rosa are present, but the scanty material obtainable did not admit of certain determi- nation. Leginuinosae 114. Trifolium pratcusc L. In the meadow ; occasional. 115. T. rcpcns L. In the meadow ; common. 116. Lathyrus palustris L. In the meadow ; frequent. 603 Oxalidaccae 117. Oxalis coruicidata L. In the meadow and open, drier places of the swamp meadow ; occasional. Euphorhiaceac 118. Acalypha virginica L. In the meadow and near various thickets; frequent. CaUitricliaccae 119. CaJJitrichc paliistris L. Very common (as also the next following species) in the reed swamp and moister parts of the swamp meadow in early- summer, when water is abundant. 120. C. heterophylla Pursh With C. palustris, the two species often entering mutually into the composition of a compact mat, the whole appearing to the naked eye as a single species. Balsaiuinaccac 121. Impaticiis biflora Walt. In moist, shady spots; occas'onal. Rhainuaceae 122. Rhanintis Frangitla L- Occurring in a thicket, west of Glencoe. The finding of this species in an established condition at Skokie Marsh has al- ready been recorded elsewhere ( Sherff, '12). Heretofore it has been frequent in cultivation, but our manuals list no place farther west than Ontario for the western limit of its established range. Vitaceae 123. Vitis vulp'ma L. In various thickets of the marsh ; frequent. Hypericaccae 124. Hypericum iiiajns (Gray) Britton In open spots in the swamp meadow ; frequent. 125. H. caiiadciisc L. In similar situations ; frequent. 604 Violaceae 126. Viola ciicullafa Ait. In the moister parts of the meadow; common. 127. V. papilionacea Pursh In similar situations ; common. Much of the material slightly d'ififerent from the typical form ; but Professor Ezra Brainerd, to whom some living specimens from Skokie Marsh were sent a year ago for cultivation in his own garden, kindly informs me that he considers them to be V. papilionacea. 128. V. conspersa Reichenb. In the meadow ; frequent. Lythraceae 129. Decodon verticillatns (L.) Ell. In the reed swamp; found at one station west of Glencoe. Its tough roots were observed in several cases to have impeded very effectively the progress, through the soil, of the stem-tuber? of Sagitfaria and Sparaaniuin. Onagraccae 130. Ludzigia polycarpa Short & Peter In the reed swamp and swamp meadow ; common. 131. /,. paliistris. (L.) Ell. In the reed swamp and swamp meadow ; very abundant. 132. Epilobium angiistifolium L. In the outer part of the meadow, west of Glencoe ; rare. 133. EpUobiiiui coloratiim Muhl. In the swamp meadow ; common. Appearing to pass by vari- ous intergradations into the next species. 134. B. adenocaiilon Haussk. With B. coloratum; common. 135. Oenothera muricata L., var. cancsccns (T. & G.) Robinson In open places' in the swamp meadow; occasional. Haloragidaceae 136. Myriophylluni heterophyllum Michx. In Skokie Stream; abundant in 191 2. 137. M. humile (Raf.) Morong In the Skokie Stream; very abundant in 191 1, but rare in 191 2, having been almost entirely replaoed by the above species. 138. Proserpinaca palustris L. In the reed swamp and swamp meadow ; very abundant. i 605 Umbelliferae 139. OsinorIii::a longistylis (Torr.) DC. About thickets; rare. 140. Cicitta macidata L. In the swamp meadow and moist parts of the meadow ; fre- quent. 141. Siiun cicutaefoUwn Schrank In the reed swamp and swamp meadow. 142. Oxypolls rigidior (L.) Coult. & Rose About thickets; somewhat rare. Cornaceae 143. Cornus Anioinum ]\IilI. In wet thickets along the marsh border, ditches, depressions, etc. ; frequent. 144. C. stolonifera Michx. With C. Amomwn; somewhat rare. 145. C. paniculata L'Her. With C. Amominn; fairly frequent. Primulaceae 146. Lysi)iiacliia thyrsidora L. In the reed swamp ; very rare. 147. Steironema ciliatum (L.) Raf. Iij shaded places in the meadow ; rare. Oleaceae 148. Fraximts amcricana L. Here and there in the small "islands" of forest. 149. Fraxinus nigra Marsh. With the last species, but in wetter soil. Gentianaceae 150. Gcntiana Andrewsii Griseb. In moist grassy thickets, west of Braeside ; rare. Asclepiadaceae 151. Asclepias incanwla L. Along Skokie Stream and in various other wet places; common. 606 Conz'olvulaceae 152. Convolvulus si'piuni L. In the swamp meadow and in the meadow ; rare. 153. Cuscuta Ccphalanthi Engehii. On Ceplialauthus occidcntalis, southwest of Ravinia; seem- ingly rare. 154. C. glouicrata Chois. On Solidago, etc., in the meadow and outer swamp meadow ; common. Vcrbeiwceae 155. Verbena hastata L. In the meadow and swamp meadow; occasional. 156. IJppici laneeolata JNIichx. In outer parts of the reed swamp, west of Highland Park; rare. Labiatae 157. Teuerium canadense L. In the swamp meadow ; rare. 158. T. occidentale Gray Mainly in the reed swamp ; abundant. 159. Scutellaria galericulata L. In the reed swamp and swamp meadow ; common. 160. Agastaclie scropluilariacfolia (Willd.) Ktze. In the meadow, near thickets and woods west of Glencoe; rare. 161. Prunella vulgaris h- In the meadow; occasional. 162. Physostegia formosior Lunell. In the reed swamp, west of Braeside; rare. The specimens do not fit descriptions of P. virginiana (L.) Benth., but match well the material collected by Dr. Lunell and described by him as new (Lunell, '08, p. 7). 163. Stachys palustris L. In the reed swamp; occasional. 164. Monarda fistulosa L. In the meadow ; occasional. 165. Pycnanthemum Hcxuosum (Walt.) BSP. In the swamp meadow and meadow ; apparently rare. fi07 i66. Lycopiis aiiii'ruaiius Mulil. Ill moist parts of the meadow, in the swamj) meadow, and along various ditches; frequent. 167. Mentha an'ciisis L., var. canadensis (L. ) Brifjuet In the reed swamp ; common. Scrophnlariaccae 168. J'crhasciini Thapsiis L. Here and there in a few open places of the meadow ; rare. 169. Chclonc glabra L. In the meadow, near thickets ; rare. 170. Mininlits ringcns L. In the reed swamp ; common. 171. Graliola zirginiaiia L. In open wet places of the swamp meadow ; frequent. 172. J'cronica Anagallis-aqiialica L. In Skokie Stream ; at many points aljundant. 173. J'. scufcHata L. In the reed swamp and swamp meadow ; frequent. 174. r. pcrcgrina L. On nude spots of soil in the swamp meadow; frequent. 175. Pedicularis lanccolata Michx. In moist places about thickets; rare. Plantayinaceac 1/6. Plantago major L. In the meadow ; occasional. 177. P. Rugclii Dene. With P. major, hut ajiparently more frequent. Rnhiaccac 178. Galiimi Claytoui Michx. In the reed swamp and swamp meadow; abundant. 179. Ccphalaiithus occiilcnfalis L. In wet thickets; occasional. Caprifoliaccae 180. Samhucus canadensis L. In thickets, mostly in outer parts of the marsh; frequent. 608 Caiiipaniilaccac iSi. Spcciilaien spots of the swam]) meadow and along ditches ; fre- quent. 202. /?. vidgata Greene In similar situations but apiiarenllv less fref|ucnt. 203. B. cennia h. In the reed swamp; abundant. 204. Hclciiiitni aiituiiinalc L. Here and there in a few moist spots; scarcely frequent. 205. Achillea Millcfoliiiiii L. In the meadow ; frequent. 206. Artciiiisia biennis Willd. In reed swamp and swani]) meadow; abundant, at least in 191 I, when marsh was fairl_\- dry. 207. Brcchtitcs liicracifolia (L.) Kaf. In a few open spots in the swamp meadow; rather rare. 208. S'cnccio aureus L. In various moist places; frequent. 209. .V. Halsaniilae Muhl. With S. aureus, or in drier soil (where more abundant). 610 210. Cirsiuiii altissiiniiiii (L.) Spreng. In open, fairly tlry places in the swamp meadow and llic meadow ; rare. 211. C. ar'c'cnsc (L-) Scop. A few fair-sized patches in drier parts of the swamp mead- ow ; scarcely frequent. 212. Taraxaciiiii officinale Weber In the meadow ; freciuent. 213. Lactuca scariola L., var. inlcgmta Gren. & Godr. In dry, open spots in the meadow and swamp meadow ; oc- casional. 214. L. canadensis L. In the meadow; frequent west of Braeside and Glencoe. 215. L. campestris Greene In the meadow ; frequent west of Braeside and Glencoe. L- campestrisy, canadensis, an evident hybrid, occurs west of Glencoe among specimens of the two species proper. The plants resemble in general appearance L. canadensis, a few of the leaves, however, becoming slightly hispid-setose underneath the mid- nerve. The corollas in the fresh specimens are a bright IMue, closely resembling those of L. campestris. Seed gathered in 191 1 was planted in 1912 but did not germinate. 216. L. spicata (Lam.) Ilitchc. Along certain ditches ; occasional. 217. Prenanthcs raceuiosa Michx. In the meadow, west of Glencoe; frequent. 611 LITERATURE CITED Atwood, W. W., and Goldlhwait, J. W. '08. Physical geography of the Exanston-Waukegan regioti. Bull. 7, 111. State Geol Surv. Baker, F. H. '10. The ecology of the Skokie ^larsh area, with .sj^cial refer- ence to the Mollusca. Bull. III. State Lab. Nat. Hist. S: No. 4. Clements, F. E. '05. Research methods in ecology. Cowles, H. C. '01. The plant societies of Chicago and vicinity. Bull. No. 2, Geogr. Soc. Chicago. Dachnowski, A. '11. The vegetation of Cranberry Island and its relation to the substratum, temperature, and evaporation. Bot. Gaz. 52 : 126-150. Flahault, Ch., and Schroter, C. '10. Phytogeographic nomenclature. Circular 6, Intern't'l Con- gress, Brussels. 28+x pp. Fuller, G. D. '11. Evaporation and plant succession. Bot. Gaz. 52:193-208. '12. Evaporation and the stratification of vegetation. Bot. Gaz. 54 : 424-426. Gluck, H. '05. Biologische Untersuchungen iilier Wasser-und Sumpfge- wachse. Erster Teil. Harshberger, J. W. '04. Phytogeographic .sketch of extreme southeastern Pennsyl- vania. Bull. Torr. Bot. Club 31:125-159. Ilenslow, G. '11. The origin of njonocotyledons from dicotyledons, through .self-adaptation to a moist or aquatic hal)it. Ann. Botany 25: 717-744- King, F: H. '97. The soil. 012 Livingston, B. E. '07. Evaporation and plant development. Plant World 10: 269-276. 'loa. A rain-correcting- atmonieter for ecological instrumentation. Plant World 13:79-82. 'lob. Operation of the porous cup atmonieter. Plant World 13: 111-119. IvUnell, J. '08. Physostegia formosior. Bull. Leeds Herb., No. 2 : p. 7. Massart, J. '03. Comment les plantes vivaces maintenent leur niveau souter- rain. Bull. Jard. Bot. I'fitat Bruxelles 14: 1 13-142. M'Nutt, W., and Fuller, G. D. '12. The range of evaporation and soil moisture in the oak- hickory forest association of Blinois. Trans. 111. Acad. Sci. 5:127-137. Pieters, A. J. '01. The plants of western Lake Erie, with observations on their distribution. Bull. V. S. Fish Comm. 21 : 57-79. Robinson, B. !<., and Fernald, M. L. '08. Gray's New Manual of Botany, 7th ed. Schouw, J- F. '22. Grundtraek til en almindelig Plante-geographie. Sherff, E. E. '12a. The vegetation of Skokie Marsh, with special reference to subterranean organs and their interrelationships. Bot. Gaz. 53:415-435- '12b. Range extensions of Rhaiiuius Praugiila and Sporoboliis aspcrifoliiis. Rhodora 14:227. 'i2c. Competition and general relationships among the subter- ranean organs of marsh plants. Trans. 111. Acad. Sci. 5: 125- 127. '13. Evaporation conditions at Skokie Marsh. Plant World 16: 154-160. Spalding, V. M. '09. Problems of local distribution in arid regions. Am. Nat. 43 : 472-486. 613 Transcau, E. N. '08. The relation of plant societies to evaporation. Bot. Gaz. 45 : 217-231. Warming, E. '09. Oecology of plants. Wilson, M. '11. Plant distribution in the woods of X. E. Kent. i. Ann. Botany 2^ : 857-902. Woodhead, T. W. '06. Oecology of woodland plants in the neighlxjrhood of Hud- dersfield. Journ. Linn. Soc. 37:333-406. Yapp, R. H. '08. Sketches of vegetation at home and abroad. IV. Wicken Fen. New Phytol. 7:61-81. '09. On stratification in the vegetation of a marsh, and its ref- lations to evaporation and temperature. Ann. Botany 23 : 275-320. EXPLANATION OF PLATES Pl.\te LXXXVI Fig. I. Map of Skokie Marsh; the dotted line represents Skokie Stream. PL.\rE LXXXVII Fig. 2. Skokie Stream at point west of Braeside, looking nortli. July, igii. Fig. 3. Skokie Stream at point west of Glencoe. looking south. Jul.v, 191 1. Pl.^te LXXXVIII Fig. 4. Average daily evaporation rates in (a) center of reed swamp, (b) outer part of reed swamp, (c) swamp meadow, and (d) forest. Fig. 5. .\verage daily evaporation rates among Pliniiniiilcs coiiniinuis: at a, o cm. : at h. 25 cm. ; at c. 107 cm, ; and at d, T08 cm. I'ig. 6. Average daily evaporation rates among T'y/'/u/ Ititifolia: at a, o cm.; at b. 25 cm.; at r 107 cm,; and at d. 175 cm, Pl.\te LXXXIX Fig. 7. a. Sl^arnanium curycarl>uiii : h. Sau.Ulnria hilifoVui: c. Pnlyi/oiiidii Muhleit- hcrgii. July, IQII. Fig. 8. Pliratj miles coiniiiuiiis. July, loU- 614 Plate XC Fig. 9, 0, Ranuncidus dcl/^hinifolius: b. Nyinl^haed adreiia; c, Siuin cicutae-i folium; d. Tyfha latifolia; c. Polygonum liydropipcroidcs. July, IQII. Fis- 10. a, Acorus Calainiis: b, Polygonum Muhloibcrgii; c, Galium CUiyfoiii. July, igii. Plate XCI Fig. 11. (7, Bolionia astcroidcs; b, Pcntlwrum scdoidcs; c. Proscr/'iiuica j'alustris; (/, Ludvigia /'a'lustris; e, Callitriche patuslris. July, 1911. Fig. 12. (I. Asclcl^ias iiicariiata: b, Poa f'ratcnsis: c, Ai/rostis alba; d, Equisctuin arvense ; e, Acalyfha Tirginicu; /, Elcocharis pahistris. July, 191 1. Plate XCII Fig. 13. a, Lycopus amcricanus; b, I'iola conspersa; c, Viola cucullata: d, Iris versicolor. July, 1911. Fig. 14. One of the broad drainage ditches southwest of Skokie Marsh (west of Kenilworth). Spargaiiiu)ii eurycarpum abundant on bed. June, 1912. Plate XCIII Fig. 15. Looking east along the county line (Braeside) road, from Skokie Stream. Swamp meadow on either side and Uliniis amcricana conspicuous in the distance. May, 1912. Fig. 16. Looking west along the road just north of the county line (liracside) road. One of the ditches. May, 1912. Plate XCIV Fig. 17. Looking north from the bridge west of Winnetka; showing a basin (largely artificial) in which part of the water from tlie marsh collects, flowing thence southward through the ditch visible in the foreground. June, 1912. Fig. 18. Looking south in tlie west part of Skokie Marsh, west of Glencoe ; show- ing the dense growth of sedges, grasses, etc., of the swamp meadow. June, 1912. Plate XCV Fig. 19. Looking south over the swamp meadow, west of Glencoe ; showing the later growth of herbs after the mowing of the tall grasses and sedges. August, 1911. Fig. 20. View west of Glencoe; the reed swamp at this point separated from forest by only about 15 ni. June. 1912. Plate XCVI Fig. 21. View in swamp meadow, west of Ravinia ; showing one of the cliarac- tcristic thickets of Salix longifolia. June, 1912. Fig. 22. Swamp meadow west of Glencoe, with Ulinus amcricana. a conspicuous tree in the landscape of the marsh, along ditch, stream, and marsh border. Tune, 1912. Pl.\te XCVII Fig. 23. View west of Highland Park, where Skokie Stream widens out but is filled with a dense growth of Sparganium eurycarpum, etc. The stream is bordered with SaVuv sp., etc. (on reader's left) and Populus Iremu- loides, Fraxinus nigra, etc. (at right). June, 1912. I'ig. 24. Skokie stream west of Glencoe, looking south; showing the numerous plants that grow rapidly upon the stream bed as the water subsides. June, 1912. Plate LXXXVI i 1 1 I Ki\on\fi(rs WVt Co\XV\\\ly Goo^ CoviTvW Fig. 1 Plate LXXXVII Fig. 2 Fig. 3 Plate LXXXVIII MAY Plate LXXXIX Fig. 7 Fig. 8 Plate XC Fig. 9 Fig. 10 Fig. 11 Fig. 12 Plate XCII Fig. 13 Fig. 14 Platr XCTTT Fig. 15 Fig. 16 Plate XCIV Fig. 17 Fig. 18 Plate XCV Fig. 19 Fio. 20 Plate XCVI Fig. 21 Fig. Plate XCVII Fig. 23 Fio. 24 I