BULLETIN
r^vToF MiJ^-

OIVIS'ON OF

NATURAL HISTORY,

Champaign, Illinois.

VOLUME II

^c?O

Article VII.—Studies of the Food of Fresh-Water

Fishes.

BY S. A. FORBES.

1888.
J. W. FRANKS & SONS, ^KINTEKS A.VD UINDEltS.

PKOBIA, ILLINOIS.



Article VII. — Studies of the Food of Fresh-Water Fish

By S. A. Forbes.

FAMILY QADID^.

Lota maculosa, Pennant. Burbot ; Lawyer.

The cod family is represented in Illinois by only a single

species, the burbot {Lota maculosa)^ occurring in the interior

of Lake Michigan, and making its way at irregular intervals to

the shallow waters within the reach of ordinary fishing opera-

tions. Since the opening of canals between the Great Lakes

and the river systems of the State, occasional specimens have

been taken in the Illinois and Mississippi.!

Its predaceous character is too well known to make

special description of its alimentary structures necessary. It

is reported by Mr. G. Brown GoodeJ to feed upon various small

fishes and Crustacea which frequent the bottom, devouring more

particularly fishes with habits like its own. It is extremely

* This article is to be considered as a continuation of the studies

reported in Volume I. of the Bulletin of the Illinois State Laboratory

of Natural History, Nos. 3 and 6, the first published in 1880, and the

second in 1883.

The data here presented relate to the fishes of the State of Illinois,

and most of them to the lower families of the series. They are derived

from collections made by my assistants and myself in various parts of

the State at intervals from 1876 to 1887, for the special purpose of

ascertaining the principal characters of the food, and the feeding

habits of the fishes of our native fauna.

tl have seen a specimen taken from the Mississippi at Canton,

Mo,, in 1887, and sent to Mr. S. P. Bartlett, one of the State Fish

Commissioners of Illinois. One occurred some years ago at Naples,

on the Illinois River, and in a letter dated April 10, 1886, Prof. J.

Lindahl, of Augustana College, Rock Island, says that three speci-

mens have been taken from the Mississippi River within his knowl-

edge, all small, the largest hardly a foot in length.

+" The Fishery Industries of the United States." p. 2.39.

1



434 Illinois State Laboratory of Natural Jlistonj.

voracious, with a wonderfully distensible stomach; and not

only captures the most active fishes, such as the pike, but will

eat carrion, and may even swallow stones. It is reported to be

nocturnal in habit, and often to secure its prey by stealth.

It is illustrated in our collection by ten examples; five

taken in spring and five in November. All but one had eaten

fishes, these making eighty-three per cent, of the food of the

entire group. One of the spring specimens had taken cray-

fishes only

—

Camharus ijropinquus^ the species commonest in

the lake. Two others of this lot had likewise eaten crayfishes,

fifty per cent, of the food of one and fifteen per cent, that of

the other consisting of this same species. The fishes taken,

with the exception of one young white-fish (
Coregonus dupei-

formis) and a small unrecognizable residue, were the common

perch of the lakes, Perca lutea.

FAMILY ESOOID^.

This family is represented within our limits by the Euro-

pean species, Esox lucius (the so-called common "pickerel" of

the streams and smaller lakes of Illinois), by the noble muskal-

lunge, Esox nohilior of Lake Michigan, and by the small grass

pickerel, Esox umbrosus. No fishes of our waters, unless it be

the gars, have become so strictly adapted to a predaceous life,

—an adaptation which probably limits them, nolens volens, to

a living prey.

Esox LUCIUS, Linn. Pike ; Pickerel.

Our specimens of this species, thirty-seven in number, of

nine different lots, were from various parts of the Illinois

River, except a single one from Fourth Lake in northern Illi-

nois.

One had eaten larvge of dragon flies (twenty per cent.), but

the entire food of the remainder consisted only of fishes, these

making, consequently, ninety-nine per cent, of the whole.

Nine per cent, were not otherwise recognizable. Twenty-one

per cent, were sunfishes and black bass—one of the latter the

small-moathed species—and nine per cent, were croppie (Po-



The Food of Fresh-Water Fishes. 435

moxys),—eaten however by only one of the specimens. Twenty
of the thirty-seven pike had taken gizzard shad (Dorosoma),

virhich made forty-six per cent, of the entire food of the species.

CyprinidsB (chiefly Notropis hudsonius) were found in two, and
three had eaten buffalo fish {Idiobus cyprinellus and /. bubalus).

Esox vERMicuLATUs, LeS. Brook Pickerel.

This fish— so far as its food structures are concerned a

miniature of the preceding—is abundant throughout the State

in ponds and lakes and along the borders of streams, especially

by the weedy margins of rivers. I have studied the food of

eighteen examples, and found it to differ from that of the larger

species only as was to be expected from the smaller size of this

pickerel, (which rarely reaches a foot in length), and from the

character of its favorite haunts.

The specimens selected for examination were from various

localities in northern, central, and southern Illinois; repre-

sented lakes, rivers, and smaller ponds; and were collected in

June, July, and October of different years.

Two had eaten the tadpoles of frogs, and eight had cap-

tured fishes,—which made about half of the food of the entire

group. Only three of these were recognizable ; one a cyprinoid,

one a sunfish, and the other (Gamhusia patruelis) a common
top minnow of the southern part of the State.

Aquatic insects formed the next most important element

of the food, reaching thirty-five per cent., and eaten by nine of

the specimens. The greater part of these were larvae of Agrion

and larger Odonata, only four per cent, being Hexagenia larvae.

One specimen had taken an isopod (Asellus), but no other

crustaceans occurred.

The food of this group may consequently be generalized

as consisting of the larger aquatic insect larvae and the smal-

ler fishes in nearly equal ratio, with occasional larva? of

Batrachia.*

Five additional specimens of this species, too large to be reck-

oned examples of the young and yet too small to class as adults had
eaten, like the full-grown examples, chiefly fishes and neuropterous
larvse. A specimen only an inch and a half in length had swallowed
a fish : one three and a fourth inches long had likewise t<kon only a



436 Illinois State Laboratorij of Natural Histortj.

FAMILY SALMON1D.E.

The common lake trout, the white-fish, and the lake her-

ring, are the representatives of this great family in the waters

of Illinois, and occur there only in Lake Michigan. None

of the smaller lakes of the State contain the herring, or so-

called '' Cisco," as do some of those of Indiana and Wisconsin.

The food of the trout and adult white-fish having been al-

ready studied by the assistants of the United States Fish Com-

missioner,* I have given them no special attention.f

COREGONUS ARTEDI, LeS. LaKE HeRRING.

Grills long, deeply arched; gill chamber consequently capa-

cious but narrow. Gill-rakers rather long and slender, allowing
.

considerable separation of the gills. Only one row on the

anterior arch, about thirty-eight in number, projecting almost

directly forward, at least equal in length to the corresponding

filaments of the gill. The anterior row on the second gill are

as stout as those of the first, but only half as long; the second

row represented by about ten triangular rudiments at the lower

end of the arch. Anterior row gradually shorter on succeeding

gills, posterior row longer; the secon drow on the fourth gill op-

posing a similar series on the pharyngeal arch. Each filament

with a double row of fine teeth along the inner edge. No phar-

yngeal teeth ;
pharynx with numerous fine longitudinal ridges

which are covered with minute recurved spines. Intestine short

and straight, anterior part provided with an immense number of

small coeca. Alimentary canal a little shorter than the head

and body without the tail.

My specimens of this species available for a study of their

food were only five in number, obtained at South Chicago in

small fish ; and a third, five inches in length, had eaten a young cen-

trarchid. The two others, respectively two and three fourths and

four inches long, had filled themselves with larvif of Agrion and

small libellulid larv.T. One had taken, iii addition, a minute larval

Corixa and a small univalve moUusk.
* See "The Fishery Industries of the United States," pp. 490, 513.

t For a discussion of the first food of the common Avhite-fish, see

Bulletin 111. St. Lab., Nat. Hist., Vol. I., No. 6, pp. 95-109.



The Food of Fresh-Water Fishes. 437

October, 1881, and at Chicago in 1885. Numbers of others

were examined, but without result, as they had been kept until

the food was all digested.

These five specimens had takeu only animal food, one of

them only Entomostraca— ninety per cent, of these being the

common Daphnia of the lakes {D. hyaJina), and the remainder

consisting of a few specimens of Bosmina, Chijdorns spheriais^

and Cyclops. The food of the remaining four was altogether

insects of terrestrial origin. In one were recognized great

quantities of winged ants (Myrmicida?), another had eaten

only Lepidoptera, and still another winged tipulids (craneflies).

In the food of one, numei'ous specimens of the common squash

beetle {Diabrotica vittafa) were recognized, and a large quan-

tity of undetermined Homoptera. An example of the homop-

terous insect Diedrocephala moUipes was detected in another.

Two small specimens of this species, hardly to be classed

as young, respectively two and six inches long, had fed, like

most of the adults examined, chiefly upon terrestrial insects,

the shortest specimen upon small Diptera (ninety per cent.)and

the homopterous insect Typhlocyba. The other example was

taken from the stomach of a lake catfish {[ctalurus lacustris)

from Lake Michigan. It had eaten a variety of terrestrial

species, including an ant, several minute Homoptera, Coriscns

ferus, a species of Amnestus, and examples of the families

Staphylinidae and Anthicidas.

FAMILY DOROSOMATID^.

DOEOSOMA CEPEDIANUM, LeS. GiZZARD SHAD ; HiCKORY ShAD ;

Mud Shad ; Thread Herring.

This remarkable fish occurs everywhere in the larger

streams and in the ponds connected with them, but not in iso-

lated lakes. It is marine in origin, swarming in the coast

waters from Delaware to Mexico.

The mouth is toothless except in youth.* The gills are

remarkably disposed within a rather small gill chamber. The

* For its juvenile characters and an account of the food of the

young, see Bulletin 111. St. Lab. Nat. Hist, Yol. I., No. 3, pp. 68-70.



438 Illinois State Laboratori/ of Natural Historij.

dorsal portion of each gill projects far forward in the palatal

region, and then turns abruptly backward, forming an acute

angle in the roof of the mouth. This course of the arches is

necessitated by the large accessory organ upon the fourth

branchial arch.* The arches are all provided with numer-
ous short rakers projecting horizontally upon either side, and

forming an unusually effective straining apparatus. The in-

testine is very long and slender and much convoluted, the

oesophagus small and long, and the stomach very short and
muscular, like the gizzard of a granivorous bird. The small

intestine is beset with a multitude of slender coeca, and its

mucous surface is everywhere remarkably villose.

The species was represented in our collections by many
specimens, but the food was so uniform in character that a

prolonged study of it seemed unnecessary, especially as the

critical analysis of such large quantities of material, minutely

divided and thoroughly intermingled, was a very tedious and

time-consuming process.

The adult specimens examined were eleven in number

;

ten from the Illinois river between Havana and Ottawa, and
one from the Pecatonica, in northern Illinois. Eight dates and
five localities are represented by them, the former ranging from
April to October.

The species has, in general, the habit of swallowing quan-

tities of fine mud, containing, on an average, about twenty per

cent, of vegetable debris. Occasionally, in the vicinity of dis-

tilleries, it feeds, like the buffalo fish, on distillery slops, and
sometimes a greater percentage of vegetation occurs mingled
with the mud. Traces of animal food were common; but the

ratio in most of ray specimens was insignificant, averaging only

four per cent, of the whole; although in one shad taken in

spring in northern Illinois one fourth of the food consisted of

Entomostraca (Cypris). Univalve mollusks occurred in one,

fragments of Coleoptera in another, and young Corixa in still

another; and spiders and water mites were also noted. Five

specimens, in all, had taken Entomostraca— four of them

* This accessory organ is correlated by Sagemehl with the 11-

mophagous habit of the fishes in which it occnra.—Morphologisches
Jahrbueh, XII., p. 318.



The Food of Fresh- Water Fishes. 439

Cypris, one Cyclops, and two Alona. The vegetable food of the

group amounted to thirty-two per cent., eaten by all the speci-

mens. Beside the distillery slops already mentioned, Lemna,

WolfBa, various diatoms and other unicellular plants, and oc-

casionally filamentous Alga3, were noted in the food. It is

probable that in some situations and at some seasons of the year,

Entomostraca would be found a more important element; other-

wise one can hardly see the advantage of the excellent bran-

chial strainer borne by this species. The great length of the

intestine and the unusual development of the mucous surface

are seemingly correlated here, as among the cyprinoids, with

the limophagous habit.

In five specimens, two and a half inches in length, the

food was intermediate in character between that of the adult

and that of the young, about sixty per cent, of it being Algai,

mixed with an abundance of dirt, and the remainder Cladocera

(twenty-two per cent.) and insect larva3— about half of them

Chironomus.

A single specimen, five and a fourth inches long, had fed

principally on Entomostraca (Bosmina, Daphnia, and Cyclops),

with a very few Chironomus larvae.

FAMILY OLUPEID^.

Only a single species of the herring family occurs in this

State— the golden shad, Clupea cJirijsochloris^ Raf.— and this not

by any means commonly with us. It seems to be strictly pre-

daceous, the three specimens taken by me at Pekin and Peoria

in September and October of three different years having eaten

only fishes— two of them the gizzard shad (Dorosoma) and

the third some undetermined kind. A single small specimen,

two and a fourth inches long, had fed wholly upon terrestrial

insects, among which were noticed Triphleps insidiosiis, aspecies

of Typhlocyba, a chalcid (Eurytoma), small Diptera (including

Culicidae and Muscidae), and some small spiders.



440 Illinois State Lahoratonj of Natural History.

FAMILY HYODONTID^.

Hyodon" tergisus, LeS. Moon Eye; Toothed Hereiistg.

This species, not common in our collections, is represented

in these studies by only five specimens, obtained from the Illinois

River at Peoria and Havana, on four dates in August and

October of two different years (1878 and 1887). Their food

consisted wholly of insects (two thirds of them terrestrial) with

the exception of a trace of univalve MoUusca. A single one,

two and seven eighths inches long, had derived its food about

equally from terrestrial and aquatic insects, including Orthop-

tera, Chironomus larvae, and Corixa tumida.

FAMILY OATOSTOMATID^.

One of the most striking characteristics of the fish fauna of

Illinois, and indeed of the Mississippi Valley, is the prominence

of the sucker family, which includes within our limits six

genera and fifteen recognized species. Several of these are

among the most abundant of our larger fishes, and most are

very generally distributed.

With reference to the essential characteristics of their food,

I find them dividing into three tolerably distinct groups. The
first includes the cylindrical suckers (Moxostoma, Catostomus,

and the like), in which the pharyngeal bones are heavy, the

lower teeth thick and strong, usually with a well-developed

grinding surface, and the gill-rakers short, thick, and few.

In the second are the deep-bodied suckers, in which the

pharyngeal jaws and teeth are well developed, although not as

strung as in the cylindrical group, while the gill-rakers are of

moderate length and number. The third contains the still

deeper-bodied and thinner species, with light pharyngeal jaws

and teeth, and long, slender, and more numerous gill-rakers.

To this group belong the species commonly placed in the

genus Carpiodes. Or, if we arrange the genera in a series, with

reference to their food structures, we shall find Placopharynx

at one extreme and Carpiodes at the other, the change consist-

ing in a gradual increase in number, length, and effectiveness



The Food of Fresh-Water Fishes. 441

of the gill-rakers, correlated with an increase in length of the

pharyngeal bones and in the number of their teeth, and a

converse diminution in the size and strength of these structures.

The intestine also becomes longer and smaller as one passes

from the cylindrical suckers to the deep-bodied buffalo and carp.

The data concerning the food of this family here presented

are drawn from a study of the alimentary contents of one

hundred and nine specimens, collected chiefly from the Illinois

and Mississippi Rivers and their immediate tributaries. They

indicate, in general, that about one tenth of the food consists

of vegetation, taken chiefly by the buffalo fishes (Ictiobus),and

in them largely composed of distillery slops. Mollusks and

insects appear in nearly equal ratio in the food of the family at

large, the former taken much the more generallv by the

cylindrical suckers, with heavy pharyngeal jaws and solid teeth,

and the latter about equally by all, with the single exception of

the stone roller ( Hypentelium ), whose peculiar haunts and feed-

ing habits explain its departure from the average. On the other

hand, the ten per cent, of Entomostraca were eaten chiefly by

the deeper-bodied species.

Placopharynx carinatus. Cope.

This species has the general appearance of one of the red

horse (Moxostoma), and has possibly been commonly over-

looked in our collections, as we have noted it very rarely.

Its branchial apparatus is not noticeably different from

that of the following genus, the gill-rakers being short and

few, and effective only on the upper part of the arch, the lower

arm being, like that of Moxostoma, covered by a ridged pad.

The fish is very remarkably distinguished, however, by the

heavy pharyngeal jaws and the thick and strong pharyngeal

teeth with conspicuous grinding surface. The latter number

about thirty on each pharyngeal, the upper ones minute and

useless rudiments, the lower ten very large, occupying about

two thirds the length of the arch,— the lower six, in fact,

one half of it. It is altogether likely that this apparatus is

related to a preference for molluscan food, but the number of

specimens available for my examination was too small to verify

this supposition.



442 Illinois Staie Laboratory of Natural Historij.

Two large examples taken from the Illinois at Havana in

October, 1887, were found to have eaten similar food. In one,

sixty per cent, consisted of small univalve Mollusca (Valvata

carinata and Amnicola), the remainder being almost wholly

insects— chiefly larvae of water beetles (HydrophilidEe) and

larval Ephemeridse (largely Caenis). About five per cent, of

Lemna occurred in this fish,— probably taken by accident, as

the river was covered with a film of duckweed at the time. A
few Chironomus larvae and an AUorchestes were also noted.

In the other specimen only five per cent, of the food consisted

of mollusks (the same species as before, together with a small.

Sphserium). Larval Hydrophilidae made eighty per cent, of the

contents of the intestine, and Ephemeridae (Caenis) more than

ten per cent. Chironomus and other dipterous larvae, Pluma-

tella, and a little Wolffia, were likewise recorded.

In a third example, only five and a half inches long, the

locality of which is not known, the food was chiefly Plumatella,

the only other elements being small case-flies (Phryganeidae), a

minute univalve shell (Strepomatidae), and a few small Chiron-

omus larvae.

MOXOSTOMA MACKOLEPIDOTUM, LeS. COMMON ReD HoRSE ;

White Sucker.

The genus Moxostoma, the commonest and most typical of

the cylindrical suckers, is represented in Illinois by three

species, two of which, aureoliim and macrolepidotum, occur

everywhere in lakes, rivers, and smaller streams. We have

encountered M. carpio but rarely, and my studies relate only to

the two former species.

In macrolepidotum the gill-rakers of the anterior row are

twenty-five to twenty-seven in number, the upper twenty to

twenty-two being elongate, triangular, stout, and crenate within,

about three fourths as long as the filaments of the gill;

while the lower five or six of this series, all of the second

series of the anterior arch, and all of the other rakers of the

gills, including the row upon the pharyngeals, have the form

of transverse leaf-like plates with crenate edges, projecting in

triangular outline a little beyond the margin of the thick gill

arch. The gills seem but slightly separable, and the branchial

apparatus is coarse and ineffective.



The Food of Fresh-Water Fishes. 448

Pharyngeals moderately heavy, the teeth about forty-five

on each side, the lower ten thickened and broadened, with

smooth terminal edges, but alternately higher and lower in the

specimen examined. The other teeth are hooked at the

anterior angle, and irregularly crenate on the cutting edge.

The intestine is small, one and a fourth times the length of

the head and body.

The salient features of the food of Moxostoma macrolep-

idotuiii, as exhibited by twelve specimens examined, are the

abundance of univalve Mollusca and the bivalve Sphaerium,

the insignificance of the vegetable element, and the absence of

Crustacea and the larger and more active insect larvas. The
insect food consisted almost wholly of larvae of Chironomus

and other small mud-inhabiting species.

The molluscan food, taken by eleven of the twelve specimens,

amounted to more than half the total, the principal forms

represented being Vivipara and Melantho (twenty-two per

cent.), Somatogyrus and Amnicola (six per cent.), and the fol-

lowing pulmonates,— Limnea, Physa, and Planorbis. Three of

the specimens had eaten Spha?rium, but the Unionidffi were

only doubtfully represented. The insects— about one third

the food— were practically all aquatic, and nearly all dipterous

larvEe. Two specimens, however, had taken a small quantity

of hydrophilid larvae, one an Agrion larva, and two others

larvffi of Ephemeridaj. The Entomostraca recognized belonged

to Alona and Cyclops. The vegetable food consisted of distil-

lery slops, eaten by one of the specimens, with a little Wolifia,

Chara, filamentous Algaj, and some miscellaneous matter.

This group of specimens was taken from the Illinois River

at Henry, Peoria, Pekin, and Havana, and from Crystal Lake

in northern Illinois, at dates ranging from May to November
of four different years.

Five additional examples of this genus, the species of

which was not determined but which almost certainly belonged

to macrolepidotum^ had eaten a still larger ratio of Mollusca

than the preceding group, these making now three fourths

of their food,— the greater part Sphaerium. Melantho. and

Amnicola also occurred, the former making one fourth of the

food of the five.



444 Illinois State Lahoratory of Natural History.

MoxosTOMA AUREOLUM, LeS. Red Hoese,

This species, less abundant in central Illinois than the

preceding, takes almost identical food, so far as one may judge

from the six specimens examined by me from Pekin, Peoria,

and Crystal Lake in northern Illinois. The food was practically

all animal, about one half of it Mollusca — largely Vivipara

and Sphgerium. The insects were, as before, mostly Chironomus

larvae and pupae, the only other form worthy of note being

smooth, slender, distinctly segmented, footless larvae with

elongate brown heads— very common in the food of fishes, but

not yet identified.*

Minytrema melakops, Raf. Striped Sucker ; Spotted Mul-
let.

In this species, not uncommon throughout the State in

suitable streams, the alimentary structures are not essentially

different from those of Moxostoma, the pharyngeal teeth being,

however, more numerous and more closely set,— about fifty-

five in the series, the lower five to ten enlarged, but less so

than in Moxostoma, and with the grinding surface less dis-

tinctly defined, most of even these largest teeth still presenting

a somewhat crenate margin.

So far as indicated by the four specimens examined, the

food of this species is similar to that of the preceding, be-

ing nearly all Mollusca,— differing, however, in the fact that

the thin-shelled bivalve Sphterium had been taken in prefer-

ence to the thick-shelled univalves. A Cyclops and a larger

percentage of Cypris represented the Entomostraca. The small

ratio of insects noticed were all Chironomus larvae.

Catostomus teres, Mitch. Common" Sucker ; White Sucker
;

Brook Sucker ; Fine-scaled Sucker.

A-bundant northward, occurring rarely in the Illinois as far

south as Peoria, and still more rarely in the extreme southern

part of the State. Wherever abundant, it inhabits nearly all

waters, both lakes and flowing streams. It is common in Lake

Michigan.

* This larva has the superficial characters of the Mycetophiliclse,

and was doubtfully assigned to that group by Dr. Williston, in a

recent letter to me.



The Food of Fresh-Water Irishes. 445

Pharyngeal jaws strong, thick, nearly twice as wide as high
;

teeth about thirty-five in number, the lower four or five much
thickened, occupying about one fourth the length of the jaw.

The crown is expanded transversely to the axis of the jaw,

rounded, not crenate or hooked. The crowns of the teeth above

the sixth or seventh are hooked and slightly crenate, but less so

than in Moxostoraa. Compared with that genus, both teeth and

jaws constitute a more effective crushing and grinding appara-

tus.

The system of gill-rakers is similar to that of Moxostoma,

but is less effective as a strainer, the anterior row of the first gill

being less numerous, shorter, and thicker. These divide into two

sets of about equal length, the upper series projecting forward,

rather short, triangular, about one third the length of the cor-

responding filaments, fifteen or sixteen in number, the lower

series, five or six, in the form of low lamellar ridges. Rakers

of the other gills thick, lamellar, with tubercles on the free

edges ; corresponding lamella" on anterior margin of the

pharyngeal jaw.

Alimentary canal about two and a half times the length of

the head and body. The alimentary structures in general in-

dicate better adaptation to molluscan food than those of the

stone roller, and inferior adaptation to Entomostraca.

The number of specimens examined was too small to make

it worth while to report their food, especially as they were

evidently under size. The branchial and pharyngeal structures

and known habits of the species indicate that its food is not

especially different from that of Moxostoma, just discussed,

and it will probably be found to consist chiefly of Mollusca

and insect larva?, the former in larger ratio than in Moxostoma,

and in smaller ratio than in the species next to follow.

Hypentelium nigricans, LeS. Stone Roller ; Hammer-

head.

This curious fish, distinguished both in form and habit

from its allies of the family, occurs usually in rapid shallows

of clear streams, commonest to the northward. It is taken

rarely in lakes.



446 Illinois State Lahorutory of Natural Histonj.

The square, strong head of this species is related to its

mode o£ life, but the cylindrical body, the large rounded pectoral

fin, and relatively high coloration, give the fish the aspect of a

darter among the suckers; and its habit of searching for its food

among the stones in swift and shallow waters is another point

of affinity with that interesting group. Curiously different as

are the food and feeding habits of this species when compared

with its nearest ally, Catostomus teres, the alimentary structures

are not remarkably unlike. The pharyngeals are somewhat

lighter, the pharyngeal teeth more slender and more promi-

nently cuspidate, and the gill-rakers somewhat stouter, possibly

affording a better apparatus for the separation of the relatively

large insect larvae upon which this species chiefly feeds. Its

alimentary structures are extremely different, however, from

those of the Etheostoraatidae, whose food, haunts, and habits

it copies so closely. It is, in short, a molluscan feeder, which

has become especially adapted to the search for insect larvae

occuring in rapid water under stones.

The pharyngeals bear about forty teeth on each side,

which are unusually high, thin, and acute, all the upper ones

with an uncommonly prominent hook or cusp at the internal

angle. The six lower teeth are cultrate, without hook or dis-

tinct grinding surface, but only two or three are noticeably

thickened.

The anterior gill-rakers are short and stout, twenty-five in

number, six of them on the horizontal part of the arch. Those

of the upper series are thin plates with the base about half the

length, and are one third to one half as long as the cor-

responding filaments. The lower rakers of the series, more
prominent than those of C. teres, are much like the upper,

but shorter, the height scarcely equal to the base. There are

five or six tubercles on the upper edge of each. The remain-

ing gill-rakers, similar to those just mentioned, interlock by

their tips, which are much more prominent and more tuber-

culate than those of Moxostoma. The stouter filaments of

the strainer are probably related to the larger and more active

insect larvas on which this species feeds.

The intestine is small, considerably convoluted, and about

twice the length of the head and body.



The Food of Fvesh-Watn- Fishes. 447

The food of six specimens taken in the Fox River and

Mackinaw Creek contained no vegetation and but a small ratio

of mollusks (Spha^rium), but was nearly all aquatic insect

larvae (ninety-two per cent.). The great majority of these

were Ephemerida?, more than half the food consisting of a

single form, abundant under stones, belonging to the genus

Caenis. A few Chironomus larvae, taken by all the specimens,

some larvae of Coleoptera, and traces of terrestrial insects were

the only other elements.

Erimyzox sucetta, Lac. Creek Fish; Chub Sucker.

Everywhere abundant in streams and lakes, ascending

creeks in spring. Occurs in our collections from McHenry to

Union county. Rarely taken by us, however, and not repre-

sented in the material used for these studies.

Pharyngeal jaws moderately heavy, short for the size of

the fish, bearing about sixty teeth, the lower ten filling the

lower third of the arch, these moderately enlarged, with incon-

spicuous grinding surface, the terminal edges being irreg-

ularly rounded. The remaining teeth are hooked, the upper

ones of the series crenate on the cutting edge.

Anterior gill-rakers thirty-four in number, upper twenty-

one short and thick, about one third the length of the gill

filaments; tips of the lower members of the series laterally flat-

tened to a paddle shape. About eight of the lower gill-rakers

of the anterior series fuse to form a thick ridged pad. Rakers

of the remaining arches similar to those of Moxostoma, but

more prominent, the tips of the transverse plates projecting

further beyond the surface of the arch.

This species presents an ovoid thickening of the palatal

region upon either side, which fills the greater part of the

branchial chamber, but is less conspicuous than in Ictiobus.

Two young specimens, one and three fourths and three

inches respectively, differed but little, in food, from those men-

tioned on page seventy-two of Bulletin 3 (Vol. I.) of the Illi-

nois State Laboratory of Natural History. The larger one had

eaten chiefly the smallest of our Entomostraca (Canthocamptus),

with a trace of Chironomus larvae. The smaller had taken a

moderate ratio of Entomostraca (Cypris, Cyclops, and undeter-



448 Illinois State Laboratory of Natural History.

rained Lynceidse), a much larger proportion of Protozoa

(especially Difflugia and Arcella), a few Squamella and other

rotifers, and unicellular Algse, including Protococcus, Chroococ-

cus, Closterium, and Cosmarium.

Genus Ictiobus. Buffalo and River Carp.

In this genus are included only the deeper-bodied suckers

with light pharyngeal jaws and relatively long gill-rakers. The

species differ, however, in these particulars, and may be arranged

in a series exhibiting a progressive lengthening of the gill

structures, a lightening of the pharyngeal jaws, and an increase

in number and a decrease in size of the pharyngeal teeth.

Related to these differences of structure are the inferior impor-

tance of mollusks in the food (especially of the thick-shelled

univalves), the greater number of insects, the appearance of

Entomostraca as an important element, and the considerable

percentage of vegetation taken. The insects eaten are well dis-

tributed instead of being essentially limited, as in Moxostoma,

to dipterous larvae. In short, correlatively with the greater

number and smaller size of the pharyngeal teeth, the weaker

jaws, and the greater development of the straining apparatus,

in Ictiobus we find the food generalized, and drawn from

numerous sources; while in Moxostoma the food and the food

prehensile structures are specialized in the direction of a rather

close dependence on the smaller mollusks.

The feeding habits of these fishes, like those of all species

inhabiting the muddy waters of central Illinois, are very diflicult

of determination, but several fishermen, and others with unusual

opportunities for observation, have reported to me that one or

more species of this genus have the peculiar habit of whirling

around in shallow water or plowing steadily along, with their

heads buried in the mud, and their tails occasionally showing

above the surface. These operations have nothing to do with

spawning, and it is likely that fishes thus engaged are burrow-

ing for small mollusks and for mud-inhabiting larvae.

Ictiobus bubalus, Raf . Quill-back ; Small-mouthed Buffalo.

This is a very abundant fish in the larger streams and in

the lakes and river bottoms, bein^ one of the three species most



The Food of Fresh- Wafer Fishes. 449

comnionl}" shipped from the lllinuis and Mississippi under the

name of buffalo fish. They all sell as "coarse fish," but from

their abundance and their fair character as food, are, on the

whole, the most important commercial fishes in our streams.

The gills of this species are very compactly disposed in a

rather small branchial chamber, the upper ends of the arches

being decurved and the lower elevated so that each gill forms

about three fourths of a circle. Ten of the lower rakers of the

anterior series are reduced to thickened ridges which extend

obliquely across the horizontal portion of the arch. The re-

mainder of this series, thirty-five in number, are flattened,

minutely toothed, the central ones about as long as the cor-

responding filaments of the gill, the others regularly shortened

above and below. The other rakers are similar to those of

Moxostoma, having the form of toothed triangular plates, with

their apices slightly projecting beyond the opposed surfaces of

the arches. The interlocking tips are a little more prominent

than in Moxostoma, and the whole apparatus is somewhat bet-

ter developed.

The pharyngeal bones are moderately heavy, triangular in

section, about as thick as high ; and the teeth, about one hun-

dred and thirty upon each jaw, project directly backwards and

act, as in Moxostoma, against a semi-circular rim of cartilage.

They are compressed, and more or less crenate on the cutting

margin, the upper ones minute, the others gradually thicken-

ing downwards so that the lower twelve occupy about one

fourth of the length of the arch. The edges of these lower

teeth are rounded, not acute.

Seventeen specimens of this species, distributed in seven

lots, collected from the central course of the Illinois River and

from the Mississippi at Quincy in the years 1880, 1882, and

1887, and in various months from April to October, give the

following general view of the food.

In decided contrast to the preceding members of the family,

about one fifth of the food consisted of vegetation— taken by

sixteen of the fishes—nearly all aquatic, but with an occasional

admixture of terrestrial rubbish. The principal vegetable ele-

ment was a small duckweed ( Wolffia ) especially abundant in

fishes taken from the Illinois during the autumn of 1887,

2



4&0 Illinois State Laboratory of Natural History.

when it made in some cases as high as ninety-tive per cent.

The larger duckweed (Lemna), fragments of Ceratophyllum,

diatoms, and other unicellular Algse, are also worthy of men-

tion.

The animal food (eighty per cent.) was fairly well divided

between Mollusca, insects, and Crustacea, respectively thirty,

twenty-nine, and twenty per cent. Only occasional traces of

univalves were noticed (Vivipara and Planorbis); but the thin-

shelled bivalve Sphgerium was a very important element, taken

by seven of the fishes, and reckoned at thirty per cent, of the

food of the group. Several individuals had eaten nothing else.

Insect larvge were very generally taken, and, in fact, oc-

curred in the food of every specimen examined. Chironomus

larvae were reckoned at nearly a fifth of the food, and were

found in fourteen out of the seventeen fishes. Neuroptera

larvae, on the other hand, occurred in relatively insignificant

number, most of them Epheraeridae ; although a small num-
ber of case-worms (Leptocerus) and of dragon-fly larvae

( A-grion) were also noticed. Hydrachnida occurred in the food

of one, and Crustacea were eaten by thirteen specimens,— all

Entomostraca with the exception of a single small crayfish

and an amphipod.

Curiously, the entomostracan eaten most freely by these

large fishes was the smallest of the Copepoda — Canthocamptus.

In the food of ten specimens taken at Peoria April 16, 1880,

and October 6, 1887, this made nineteen per cent, of the food

of the entire group. Specimens of Cyclops, Cypris, Pleuroxus,

Iliocryptus, Bosmina, and Simocephalus occurred in numbers
too small to figure in the ratios. Fresh-water Vermes were

almost wholly wanting, only a few Anguillulidae occurring in

the food of one. Eight had eaten Polyzoa, including both

Plumatella and Pectinatella. The latter was recognized by

its statoblasts only, detected in seven specimens collected in

October, 1887, in situations where the gigantic colonies formed

by this polyzoan had been earlier very abundant. It is proba-

ble, consequently, that these statoblasts, widely dispersed with

the death and decay of the translucent mass in which they are

developed, had been picked up by accident with the other food.*

* Some notes on the young of this genus, published in the Bulletin

of this Laboratory, Vol. I, No. 3, page 7.S, show that specimens varying



The Food of Fresh-Water Fishes. 451

ICTIOBUS CYPEIXELLUS, C. & V. ReD-MOUTH BuFFALO.

The statements made coueerning the abundance, distri-

bution, and commercial value of the preceding species will ap-

ply equally well to this. The fishermen report, however,

that the quill-back frequents deeper water than the red-mouth.

The structures of food prehension differ from those of bubalus

in the lighter pharyngeal jaws, the greater number and smaller

size of the teeth, and the more efficient branchial apparatus.

The pharyngeal jaws are relatively thin, the thickness being

about one fourth the height. The teeth are about seventy-five

in number on each jaw, minute above, gradually but not greatly

thickened below, the ten lowest occupying nearly one fifth the

length of the jaw. These largest teeth have the cutting edges

obtuse, and are slightly hooked within. The remaining teeth

are more or less crenate on the cutting edge, each with con-

spicuous hook or cusp at the inner angle. The posterior edges

are also acute.

The gill-rakers are similar to those of the quill-back, but

more efficient as a straining apparatus. The longer rakers of

the anterior row (seventy-five in number) are fully equal in

length to the corresponding filaments, and are armed within

with a double row of clusters of minute teeth. Eight or ten of

the lower rakers are fused in the form of thick oblique ridges,

The tips of the rakers of the other rows project beyond the

borders of the arches a distance about equal to the line of at-

tachment to the arch. The pharyngeal enlargements are very

conspicuous and thick, nearly filling the pharyngeal cavity.

in length from seven eighths of an inch to two inches, fed largely

upon unicellular Algje and rotifers, the remainder of their food being

chiefly the smallest Entomostraca. I add here the details from two

additional specimens, taken in June, from the Illinois River, at

Pekin, one three fourths of an inch in length and the other eight

tenths. The greater part of the food of these consisted of rotifers.

Protozoa, and gelatinous and other unicellular Algje, a single Bos-

mina in each being the only entomostracan form determined. The

rotifers included Brachionus and Anurea ; and among the Protozoa

were Actinosphcerium, Arcella vulgaris, and .4. discoidea. Closterium

was noticed among the Algse, with numerous gelatinous Algse related

to Protococcus, and a filament of OsciUatoria. Spores of fungi were

found in both, and a fragment of vegetation penetrated by a fungus

mvcelium occurred in one.



452 Illinois State Lahoratori/ of Natural Ilistori/.

This species seems to differ in food from the preceding,

especially in the inferior amount of mollusks and the larger

ratio of vegetation. The animal food of seventeen specimens

collected in seven lots from the Illinois and Mississippi Rivers

and the northern lakes in various months from April to Octo-

ber of five different years, was about two thirds the whole, the

remaining third consisting largely of Algae, unicellular and

filamentous, and otherwise chiefly of distillery slops (taken by

Illinois River specimens) and miscellaneous vegetation of ter-

restrial origin. This last was occasionally found in quantities

sufficient to show that it had been intentionally swallowed,

making in one instance the greater part of the food. The

molluscaa food of these specimens amounted to only three per

cent., nearly all Spheerium ; the insect food to thirty-three per

cent., practically all aquatic, and very largely larvse of Chi-

ronomus (twenty per cent.). The Neuroptera were chiefly

Hexagenia larvae (nine per cent.). Except a single Crangonyx,

the Crustacea were all Entomostraca. These occurred in

much greater variety than in cyprinellns, among them being

representatives of Daphnella,Simocephalus, Bosmina, Chydorus,

Pleuroxus, Alona, Cypris, Cyclops, and Canthocamptus. Frag-

ments of Plumatella were noticed in a single specimen, Dif-

flugia in two.

IcTioBUS URUS, Ag. Black Buffalo ; Mongrel Buffalo;

Big-mouthed Buffalo; Chuckle-head.

This species occurs commonly with the preceding, but less

abundantly. Said by fishermen to frequent shallower water.

With respect to food, it closely resembles cyprinellus, our

seventeen specimens, well distributed as to date and place,

having taken almost identical ratios of animal and vegetable

food— sixty-seven per cent, and thirty-three per cent, respec-

tively. Twelve per cent, were mollusks,— nearly all Sphaerium,

as before. The large ratio of insect food (about forty-two

per cent.) was more than half Chironomus larvae, most of the

remainder being Hexagenia larvae, taken, however, by only one

of the specimens. The Crustacea (thirteen per cent.) were

practically all Entomostraca, fragments of a young crayfish

appearing in only a single specimen.



The Food of Fresh-Wafer Fishes. 453

Among the vegetable elements, distillery slops (eaten by

three of the specimens) were the most important (twenty-one

per cent.). The rather insignificant amount of aquatic vegeta-

tion (six per cent.) was distributed as usual among a number

of the lower plants, chiefly duckweeds and the unicellular Alga?.

IcTioBus CYPRiJfUS, LeS. River Carp; Carp Sucker.

Under this specitic head I include, for the purposes of this

paper, all the so-called species of river carp sometimes separated

under the genus Carpiodes, and hitherto described under some

eight specific names. This form is abundant in the great rivers

of the State and in their larger tributaries, and also in Lake

Michigan and the smaller lakes of northern Illinois. It is

extremely common in the lakes and ponds of the river bottoms,

but occurs in running water in smaller numbers than the

other species of its genus.

In its structures of food prehension it exhibits an extreme

development and a correlative degradation of branchial appa-

ratus and pharyngeal structures respectively. The gills are re-

markably compacted, the upper and lower ends nearly meeting

when the mouth is closed. The pharyngeal protuberances are

enormous, almost filling the branchial cavity. Anterior gill-

rakers in two series, as usual, the upper about sixty-seven in

number on three fourths of the arch, the longest a little longer

than the corresponding filaments. The lower part of the gill

with about ten thick, papillar, coherent ridges extending down-

ward a distance equal to the length of the filaments of the

same vicinity. The longer rakers have each two closely alter-

nating rows of tubercles on the inner edge, roughened with

extremely minute denticles. Inner surface of the arch with

transverse tuberculate ridges springing from the bases of the

rakers of the gill, and terminating inwardly in slight projec-

tions representing the posterior row of rakers. The other

arches are similarly tuberculate and ridged, and the whole ap-

paratus closely embraces the pharyngeal thickenings. Pharyn-

geal bones very thin and brittle, less than a millimeter thick in

a fish ten inches long, the thickness one seventh the height to

the base of the teeth. The latter about two hundred, minute

above, gradually increasing downwards, but not much thick-



454 Illinois State Lahoratorij of Natural History.

ened or elongate, about thirty on the lower fourth of the arch.

Crowns emarginate or doubly emarginate, with the inner angle

similarly produced, forming a hook or cusp. Intestine very

slender, four times as long as head and body in the specimen

examined.

Nineteen examples of the species, representing thirteen

dates and localities, from April to October, and from 1877 to

1887, collected from Crystal Lake in northern Illinois, from

the lakes of the Ohio near Cairo, and from the Illinois River at

Ottawa, Peoria, and Havana, show that the native carp differs

from the other species of Ictiobus chiefly in the inferior amount

of vegetation eaten, in the greater quantity of mud mingled

with the food, in the absence of the larger insect larvae, and in

the lack of univalve Mollusca. It resembles closely Ictiobus

cyprinellus, but from this differs also with respect to the vegeta-

tion taken, and in its filthy feeding habits. The vegetable food

was only eight per cent., mostly Wolffia, and that eaten by

only two of the specimens. A few diatoms were mingled with

the mud in three, and miscellaneous aquatic vegetation occurred

in five. Mollusks made about a fourth of the food,— all the

thin-shelled Sphaerium. Insects averaged about one third, the

greater part Chironomus larvas. Neuroptera were eaten by

only four of the specimens, and contributed only two per cent,

to the food, case-worms (Phryganeidfe) being the only forms

identified. Entomostraca made nearly a fourth, distributed

through a considerable list, which included Simocephalus

americanus, Bosmina, Chydorus, Alona, Cypris, Cyclops, and

Canthocamptus. No Vermes or Polyzoa were observed, but

occasional Protozoa were noticed, especially Centropyxis and

Difflugia.

Looking now at the food of the family, as exhibited by the

one hundred and seven specimens discussed, representing, as

they do, five genera and eleven species, we conclude that the

sucker family is essentially carnivorous, the vegetable food

amounting to only eight per cent, of the whole, and no ele-

ment of this being especially prominent. The smaller mol-

lusks are the most important single class, the ratio of these

being forty-one per cent., about three fourths of them Sphae-

rium. The large quantity of aquatic insects (one third of



The Food of Fresh-Water Fishes. 455

them Chironoraus and a fourth ephemerid larvae), the relative

insignificance of Crustacea (about ten per cent.,— nearly all

Entomostraca), and the practical absence of Vermes and Pro-

tozoa are the remaining salient features of the food characters

of this family.

FAMILY SILURID^.

The family of catiishes taken together is nearly omnivorous

in habit, and their alimentary structures have a correspondingly

generalized character. The capacious mouth, wide oesophagus,

and short broad stomach, admit objects of relatively large size

and of nearly every shape; the jaws, each armed with a broad

pad of fine sharp teeth, are well calculated to grasp and hold

soft bodies as well as hard; the gill-rakers are of average

number and development; and the pharyngeal jaws— broad,

stout arches below and oval pads above, with thin opposed

surfaces covered with minute, pointed denticles— serve fairly

well to crush the crusts of insects and the shells of the smaller

mollusks and to squeeze and grind the vegetable objects which

appear in the food. The use made of the jaws in tearing mol-

lusks from their shells, as described further on, is probably the

most peculiar feeding practice of these animals; and the indif-

ference of several of the species to the past history or the present

condition of their food, distinguishes them as the only habitual

scavengers among our common fishes.

The family is a very abundant and characteristic one in

this region. It ranges in size from the smaller species of

Noturus, only an inch or two in length, to monsters more than

two hundred pounds in weight; and inhabits every kind of water

from the greatest rivers of the continent to small temporary

ponds of surface water, where its presence is the staiiding won-

der of the fisherman and the naturalist.

In Illinois we have three genera and twelve species of these

fishes, as at present classified, none of them unfit for food

except the smallest ones, and two or three of them the equals

of any river fish.



456 Illinois State Laboratory of Natural History.

My studies of their food were based upon one hundred and

twenty specimens, belonging mostly to five species of Ictalurus

and Noturus. The data are especially deficient with respect to

the food of the largest lake and river species.

ICTALUKUS FURCATUS, C. & V. CHANNEL CaT ; FoRK-TAILED

Cat; White Fulton.

This is the catfish par excellence., and is the best food fish

of its family. It occurs only in the deeper water of the larger

streams. It is common in the Mississippi and the Ohio, al-

though much less so than the following species, but is taken

rather rarely in the Illinois, where it is often called the

"Mississippi cat." It is never found in lakes and ponds, and

feeds, according to the reports of fishermen, almost exclu-

sively upon other fishes. A single specimen taken at Quincy

Oct. 25, 1887, had eaten fishes only.

The gill apparatus is better developed than in Amiurus, but

is nevertheless very incomplete. The anterior arch has only one

row of rakers, eleven in number below the angle, four or five

above. These are longest near the upper end of the lower part of

the gill, where they are about half the length of the corre-

sponding filaments. The other gills have similar but shorter

rakers, the third and fourth a double row of about equal length.

None of the rakers are toothed or tuberculate. The pharyn-

geals, both superior and inferior, are similar to those of Amiu-
rus, but relatively smaller.

Ictalurus punctatus, Raf. Blue Fulton; Spotted Cat;

Fiddler; Switch Tail.

An abundant species in the larger rivers, much commoner
than the preceding, but not quite so good for food, smaller,

ranging more freely, and clearly a more general feeder,

although its alimentary structures are not noticeably different.

The gill-rakers of the anterior arch are a trifle shorter, the

longer ones being about one third the length of the corre-

sponding filaments, and the pharyngeal structures seemingly a

little heavier.



The Food of Fvesli-Watcr Fishes. 457

Forty-three specimens of this species were taken from the

Illinois River at Peoria, Pekin, and Havana, and from the Mis-

sissippi River, near Quincj, Their dates of capture represent

the spring, summer, and autumn months of the years 1878, 1880,

and 1887.

About a fourth of the food consisted of vegetable matter,

much of it miscellaneous and accidental, but chiefly Algae—
Cladophora being the most abundant form. This and other

filamentous Algae made a large part of the food of several

fishes taken in October, 1878 and 1887, three having eaten

nothing else. Fragments of Potamogeton were taken by other

October specimens, making twenty per cent, of the food of

three. The fact that the floating Lemna occurred but rarely,

and then in the smallest quantity, is evidence that these cat-

fishes are strictly bottom feeders. A single specimen had

fed on still-house slops, as shown by the considerable amount

of meal in its alimentary contents.

A dead rat, pieces of ham, and other animal debris attest

the easy-going appetite of this thrifty species.

Fragments of fishes were found in eleven examples of this

group,— commonly, however, in pieces so large as to make it

certain that they were derived from those already dead. Occa-

sionally, as in examples taken in August, 1887, from the Mis-

sissippi River, fishes probably taken alive composed the whole

of the food. The species were not identifiable.

Molluscan food was a decidedly important element, being

found in fifteen of the fishes and amounting to fifteen per cent,

of the whole. Several specimens had taken little or nothing else,

—notably six secured at Havana in September, 1887, and one

at Peoria in October of the same year. The Mollusca were

about equally divided between gasteropods and lamelli-

branchs, the former largely Melantho and Yivipara, the latter

usually Unio or Anodonta.

Notwithstanding the number of bivalves eaten by these

fishes, no fragment of a shell was ever found in their stomachs,

but the bodies of the animals had invariably been torn from the

shell while yet living— as shown both by the fresh condition

of the recently ingested specimens and likewise by the fact that

the adductor muscles were scarcely ever present in the frag-



458 Illinois State Lahoratovij of Safiiral History.

ments. Indeed in only a single instance had the posterior ad-

ductor been torn loose. The Unionidaj were usually large and

thin — probably in most cases Anodonta.

I have been repeatedly assured by fishermen that the cat-

fish seizes the foot of the mollusk while the latter is extended

from the shell, and tears the animal loose by vigorously jerking

and rubbing it about. One intelligent fisherman informed me
that he was often first notified of the presence of catfishes in

his seine, in making a haul, by seeing the fragments of clams

floating on the surface, disgorged by the struggling captives.

Still more interesting and curious was the fact that the

univalve Mollusca found in the stomachs of these fishes were

almost invariably naked, the more or less mutilated bodies

having only the opercles attached. How these fishes manage
to separate moUusks like Melantho and Vivipara from the shell,

I am scarcely able to imagine, unless they have the power

to crack the shells in their jaws as a boy would nuts, and then

to pick out the body afterward. Certainly the shells are not

swallowed, either whole or broken.

The number of mollusks sometimes taken by a single cat-

fish is surprising. As high as one hundred and twenty bodies

and opercles of Melantho and Vivipara were counted in a spot-

ted catfish taken at Havana in September of last year.

Insects were, however, the principal food of the specimens

studied, making forty-four per cent, of all, eaten by twenty-

eight of the specimens; five, in fact, had eaten nothing else,

and nine others had taken ninety per cent, or more of insects.

These were mostly aquatic, although now and then a fish had
filled itself with terrestrial specimens. About half the in-

sects were Neuroptera, nearly equally dragon-fly larvse and

larvae of Epheraeridge; but Hexagenia larvae were rarely recog-

nized. Chironomus larvae made thirteen per cent, of the food,

aud were so frequently taken with the sand tubes they inhabit

as to make it certain that they were commonly obtained from

the bottom. Leeches appeared in the food of three of the speci-

mens, and Gordius in one. Fragments of Plumatella were no-

ticed in two, and a fresh water sponge likewise in two.

Four immature examples of this species, ranging from two
and a half to four inches in length, had fed almost wholly



Tlie Food of Fye^h-Wafer Fishes. 459

upon insects, a few specimens of Allorchestes dentata and

Daphnia being the only other items. Eggs and young of

Hexagenia and other ephemerids composed the greater part of

the food, Chironomus larvae amounting to about one half as

much.*

ICTALURUS NATALIS, LeS. YeLLOW CaT.

This species occurs everywhere throughout Illinois, but

less abundantly than )iebuIosus, and usually in larger streams.

It has not been taken by us from ponds and lakes except where

these were immediately connected with rivers subject to over-

flow.

The alimentary structures of this species closely resemble

those of /. nebulosiis, described under the next head. Twelve

specimens were collected from the Illinois River at Peoria, the

Fox River at McHenry, and from one of the smaller lakes in

northern Illinois, in the months of May, August, October, and

November of 1878, 1880, and 1887.

The food was wholly animal with the exception of a trace

of duckweeds (Lemna and Wolffia) taken by a single specimen.

The scavenger habit of the species was shown by the food of

the Fox River specimen, three fourths of which consisted of

the remnants of a dead cat. Fishes made a larger ratio of the

food than in the preceding species, amounting to about one

third, most of them apparently taken alive. One, however, a

sucker, was represented only by the stomach and intestines,

doubtless picked up near a fish boat. The gizzard shad, cer-

tain Cyprinidse, and undetermined suckers (Catostomatidae) were

recognized, four of the twelve specimens having fed wholly or

almost wholly upon them. The molluscan food of these speci-

mens was insignificant, no bivalve mollusks having been taken

by them and only a few Vivipara and Melantho, amounting in

all to five per cent. While insects had been eaten by four of

the specimens and reached a ratio of thirty per cent., they were

practically all Hexagenia larvae, taken in October, 1878 and

1887. On the other hand, seventeen per cent, of the food was

catfishes, taken by four of the specimens in May and August.

* A hmt of the winter food is given by six specimens received

from the Tllinois "River at Havana, February, 1888, all of which had

fed uiily uiion Chiiouuums huv:v or laiva' of Agrion.



460 Illinois Stciie Lahoratorij of Natural Hisionj.

Seven immature examples, from two to three and a half

inches long, had fed chiefly upon Entomostraca, which made

about one half the food. Among these, Daphnia, Simocephalus

americanus^ Acroperus, MacrotJirix laticornis^ Cyclops, and

Cypris were determined. One fourth the food consisted of the

univalve Physa, and one fifth of it of insect larvae, chiefly

epheraerids and Chironomus. A little Wolffia and other

aquatic vegetation likewise occurred.

lOTALUEUS NEBULOSUS, LcS, BULL-HEAD ; HoRNED-POUT.

This superabundant species occurs in all waters and in all

parts of the state, but frequents by preference ponds and muddy

streams. It grows to a larger size in the rivers than elsewhere,

and has many marked varieties. Its feeding habits are appar-

ently essentially the same in all situations.

Gill-rakers fourteen in number on the anterior gill, in one

row, thick, stout, not toothed, at the angle of the arch about

half as long as the filaments, shortening rapidly above and be-

low. Second gill also with a single row, shorter than those of

the first; succeeding gill with two rows each of still shorter

rakers, the posterior row shorter than the anterior ; a smaller

row upon the pharyngeal arch. The upper pharyngeals are

large and broad oval pads, with convex surfaces paved with

close-set, minute, sharp teeth, and act against the broad lower

pharyngeals, which are similarly armed. Intestine to head

and body as 1.2 to 1.

Thirty-six specimens were collected for a study of the food,

— at Normal, Peoria, Pekin, and Havana, in Central Illinois
;

and from the Fox River and several of the small lakes in the

northern part of the state. The collections were made in May,

July, August, September, and October, of four different years.

The vegetable food nearly equaled that of I. pundatus,

and was taken by seven of the specimens. One had eaten dis-

tillery slops, and in the food of the others were found Cerato-

phyllum, Potamogeton, Chara, and various Algffi.

Fishes made one fifth of the food,— taken however by only

^wo of the specimens, which had eaten nothing else. One of

the fishes was a perch and the other a sunfish (Centrarchidge).



The Food of Fresh-Water Fishes. 461

Mollusks made one fifth of the entire amount of the food,

— more than one half of them Sphaeriiim. This genus made
nearly all the food of a large group taken from the Illinois River

at Pekin in September, 1882, and also of two other specimens

taken in the Illinois River at Peoria in October, 1887. Uni-

valves were rarely preseut, amounting to only two per cent,

of the food, taken however by eight of the specimens. These

included the usual forms—Valvata, Melantho, and Amnicola,

together with two or three specimens of Physa. Examples of

Pisidiura were rarely noted, and two had eaten Unios.

Nearly a fourth of the food was insects, mostly aquatic,

and the larger part of them larvae of Diptera— especially Chi-

ronomus and Corethra. Seven per cent, of Neuroptera larvai

(Hexagenia, Libellulid^, and Phryganeidae), together with a

miscellaneous assortment of terrestrial species, complete the ac-

count of the insect food.

The Crustaceans (thirteen per cent.) were nearly all cray-

fish, traces of Diaptomus, Leptodora, Chydorus, etc., appearing,

however, in here and there a specimen, and the little am-
phipod AUorchesfes dentata, appearing in three. A leech and

a nematoid worm occurred, each in one.

It will be seen that the food of this species was very widely

distributed, being composed about equally of fishes, mollusks,

aquatic insects, and vegetable structures, with a very considera-

ble ratio (thirteen per cent.) of crustaceans added.

Two smaller specimens, two and three and a half inches

respectively, had fed chiefly on ephemerid and Chironomus

larvai, small crayfish, and Asellus. To these were added Corixa

tumida, Cyclops, Daphnia, filaments of Spongilla, Chydorus,

Scapholeberis mucronatus. a few Diatoms, and traces of fila-

mentous Algse.

ICTALURUS MARMORATUS, Holbrook. MaRBLED CaT.

This species is scarcely more than a deep-water variety of

the common bull-head (/. nebidosKs), distinguished only by the

color. It occurs in the larger rivers of the State and their im-

mediate tributaries, but nowhere, so far as I know, in stagnant

waters. Our thirteen specimens were all from Peoria and

Havana, taken in August, October, and November of 1878 and

1887.



462 Illinois State Laboratory of Natural History.

The food of this species as represented by these thirteen

specimens, is unusually simple for a catfish, consisting chiefly

of bivalve mollusks, larvae of Chironomus and Hexagenia, dis-

tillery slops, and accidental rubbish. Fishes are conspicuous

by their absence, only a single specimen exhibiting any trace

of them.

Sphaarium and Unio made about a fourth of the food, and

aquatic insect larvae amounted to one half (Hexagenia thirty-

five per cent, and Chironomus fourteen). A hydrophilid and

a few terrestrial insects, a few specimens of Vivipara and a

Physa, sialid larvae (taken by two), slender leeches eaten by

five, and a trace of Potamogeton in one, are the minor ele-

ments of this record. One of the specimens, taken in Novem-

ber, had eaten eighteen leeches, which made one fourth of its

food. It will be noticed that three fourths of the food con-

sisted of bivalve mollusks and insect larvae.

Leptops olivaris, Raf. Mud Cat; Yellow Cat; Morgan Cat.

Common in the deeper waters of the larger streams. Ob-

tained by us only from the Illinois, Wabash, and Ohio.

This is one of the largest of the river catfishes, repulsive

in appearance, but above the average as food. It is reported

by fishermen to feed only upon animal food— chiefly fishes—
and such was the case with the two specimens examined from

collections made at Quincy in August, 1887. These had fed

upon the common river sunfish (Lepomis), several cyprinoids,

and an Amiurus four inches long.

NoTURUS GYRiNUs, Mitch.

This little catfish, the most abundant of the small species

of the family, occurs throughout Illinois, but has been con-

fined in our collections mostly to lakes, rivers, and large creeks.

It is not by any means restricted to rocky situations, but seems

rather to prefer the muddy parts of both the rivers and lakes

in which it occurs.

Thirteen specimens were secured at Pekin and Peoria,

from Clear Lake in Kentucky, and from the Fox River in

McHenry county. Their food was wholly animal, with the

exception of a trace of Algae found in two. This group



The Food of FroHh-Water Fishes. 463

had eaten praetieallj nothing but Crustacea, nearly all Am-
phipoda ( A.llorchestes) and Isopoda (Asellus), the former
eaten by nine, and the latter by two— both together mak-
ing forty-seven per cent, of the entire food. As might be

supposed from the small size of these specimens, Entomostraca
were apparent in the food, although in moderate numbers
(five per cent). The forms recognized were Simocephalus,

Chydorus, Pleuroxus, Alona, Cypris, Candona, Cyclops, and Can-

thocamptus. A planarian worm was noted in one, and speci-

mens of Difflugia in another. A single example had eaten a

small fish. Most of the insects were Chironomus larvae (twen-

ty-five per cent.), case-worms, and larvae of day flies (twelve per

cent.).

Comparing the principal genera of this family, as repre-

sented by the one hundred and twenty specimens examined,

we find that the larger deep-water species from the great rivers

of the State are apparently ichthyophagous ; that the relatively

minute stone cats feed on the smaller insect larvee and the

medium sized Crustacea ; that the spotted cat is essentially in-

sectivorous ; that among the ball-heads the yellow cat eats the

largest percentage of fishes and the marbled cat the smallest
;

that the latter feeds more generally upon Unio than any of the

other species; and that moUusks at large make about one sixth

of the food of the group of species which feeds upon them.

FAMILY AMIID^.

Amia calva, Linn. Dog Fish; Mud Fish; Grindle.

This species is very abundant throughout the State in the

lakes and larger streams, and also common in ponds of south-

ern Illinois. Not commonly eaten, but often caught for sport.

The food of twenty-one specimens taken from northern,

central, and southern Illinois, in April, May, June, August, Sep-

tember, and October, was wholly animal, about one third of it

fishes, among which were recognized some undetermined cypri-

noids and a small buffalo fish (Ictiobus). The other important

elements were mollusks— about one fourth — and crustaceans

(forty per cent.), insects being represented by an insignificant

ratio (two per cent.). Even the usually abundant Chironomus



464 Illinois State Lahoratory of Natural History.

and ephemerid larvcB had been eaten by only one or two

specimens each. The mollusks were more than two thirds

Sphseriurn, the remainder being Vivipara and Planorbis. The

Crustacea were chiefly crayfish, among them Cambarus virilis

and obesus. Besides these, I noticed Crangonyx, Allorchestes,

and Asellus, Cyclops and a few Cladocera (Simocephalus ameri-

canus^i Scapholeberis mucronatus, and Chydorus).

FAMILY LEPIDOSTEIDuE.

A half dozen of the river gars Lepidosteus platystomus

and L. osseus had eaten nothing but fishes, including the hick-

ory shad (Dorosoma), black bass (Micropterus), and some min-

nows (Cyprinidae).

FAMILY POLYODONTID^.

PoLYODON sPATHULA, Wall. Shovel-fish; Paddle -fish;

Spoon-bill Cat; Duck-bill Cat.

This remarkable and most interesting fish, the most nota-

ble inhabitant of our waters, occurs abundantly in the Illinois,

Mississippi, and Ohio, but not elsewhere within our limits. It

has a more or less distinct habit of migration, being much the

most abundant in spring, although taken sparingly throughout

the remainder of the year. It is a gigantic species, reaching a

weight of thirty pounds and upwards, and a length of six feet

or more, including the paddle. It is now quite generally

dressed for the market, and sold at the same rate as catfish.

It has an alimentary apparatus not less remarkable than

its other characters. The broad blade-like snout, the enormous

mouth and equally large gill slits, the efiicient branchial

strainer, and the peculiar structure of the intestine,— all in-

dicate a peculiar alimentary regimen and unusual feediug

habits. Both the upper and lower jaws of the young are pro-

vided with small, acute teeth— the upper with a band upon

the vomer and palatines, besides a row on the maxillaries, and

the lower with a longitudinal row extending nearly its full

length— but the jaws of the adult are toothless and smooth.



The Food of Fresh-Water Fishes. 465

This fish depends, therefore, eutirely upon the very remarkable

straining apparatus borne by the gills, the iramense oral open-

ing, and the equally free provison for the exit of water from

the gill chamber, enabling it to pass vast quantities of water

through its branchial apparatus. The gills are very elongate,

each having the form, when the mouth is closed, of a slender U
with the sides parallel and closely approxin ated, the lower arm,

however, extending soniewhat turliier forward than the ufjper.

Each gill bears throughout its whole length a double series of

very long, fine, numerous, and slender rakers, the two rows

separated by a membranous partition borne upon the anterior

surface of the arch,— this partition a little higher than the rows

of rakers, and slightly thickened on the internal edge, so as to

enclose the tips of the rakers when the parts of the apparatus

are approximated. These rakers average fully twice the length

of the corresponding gill filaments, and numbered, on the first

gill of a specimen about one and a half feet long, five hundred

and sixty rakers in the anterior series. A half row of similar

rakers is borne by the fifth branchial arch, corresponding to

the inferior pharyngeal bones of most fishes. The individual

rakers are toothless, smooth, cartilaginous, and nearly naked,

the filaments covered by a thin epithelium, thickened at the tip.

Interlocking as these do when the branchial apparatus is ex-

tended, they form a strainer, suJBBcient to arrest the smallest liv-

ing forms above the Protozoa. There are no pharyngeal jaws

or teeth, nor is there any apparatus of mastication elsewhere.

In the absence of any raptatorial teeth or crushing appa-

ratus in its large and feeble jaws or in its throat, it is certain

that this species cannot feed upon fishes or moUusks ; and

the character of the intestine makes it very probable that it

never purposely swallows mud or takes a large percentage of

vegetable food. On the other hand, its enormous mouth, and

the remarkable straining apparatus in its branchial cavity give

it access to the immense stores of minute insect and crustacean

life most commonly reserved for young fishes ; while its struc-

tures are likewise evidently adapted to the larger soft-bodied

insects and insect larvtc.

The use of the paddle-like snout is as yet a matter of con-

jecture, slightly assisted, perhaps, by a knowledge of the princi-



466 Illinois State Laboratory of Natural History.

pal features of the food. The relatively minute size of the ob-

jects on which it feeds, the absence of mud from its intestine,

and its seemingly positive preference for animal food, indicate

that it is not only able to gather large quantities of very mi-

nute objects among the weeds and from the muddy bottom

without filling itself with mud, but that it can separate the

Entomostraca from the Algge among which they swim. I

cannot see how this is done unless its paddle be used to stir up

the weeds in its advance, as it swims along, thus driving up the

animal forms within reach of its branchial strainer, while the

mud and vegetation settle out of its way.

What is the meaning of the minute and evanescent teeth

on the jaws of Polyodon, I am unable to surmise, but judge

that they can only be accounted for by reference to primitive

conditions of life of which the present habits of the fish give

us no hint.

Eight specimens obtained from Peoria, Pekin, and Henry
on the Illinois, from the Ohio River at Cairo, and from the

Mississippi at Quincy, in six different years, will probably suffice

to give a fair general idea of the food, taken in connection with

suggestions made above, based on a study of the structures of

alimentation.

The vegetable elements of the food were eaten by four of

the specimens, and amounted to only seven per cent. It is to be

noted, however, that one of the specimens taken at Quincy

had derived thirty per cent, of its food from a species of Nostoc,

while another, taken at Peoria in May, had found about one

fifth of its food among vegetable objects. A little Potamo-
geton, some filamentous algse and diatoms, together with a small

amount of terrestrial rubbish, were the elements recorded.

Fishes and mollusks were without representation in the

alimentary contents of these specimens; while insects and crus-

taceans made by far the larger part of the food,— the former

taken by all the specimens, and in nearly twice the ratio of the

latter. The minor items of this class were Corethra larvae

(twelve per cent.) and Chironomus larvae (five per cent.).

Larvae of Neuroptera made one half the food, and were eaten

by six of the specimens,— Hexagenia larvae alone amounting

to forty-seven per cent. A. few case-worms (Phryganeidae),



The Food of Fresh-Water Fishes. 467

dragon-fly larva3 (Libellulida3 and Agrion), and Ctenis larvae,

with a few Corisas, aquatic beetles (Coptotomus), and chance

terrestrial insects, were the remaining items of this class.

The crustaceans were all Entomostraca, with the exception

of the amphipod Allorchestes dentata, noted in two specimens.

Five of the specimens had eaten Entomostraca, one of them

ninety per cent., and another eighty,— the remaining ratios

being thirty-five, thirty, and twenty. Water mites (Hydrach-

nida) were noticed in a single specimen, leeches also in one,

and Plumatella in another. The smaller Crustacea were so

numerous that no attempt was made to exhaust the possible

determinations ; but in some cursory examination of this ma-

terial the foUowiug forms were observed : Daphnki jjulex,

Bosmina, Cbydorus, Eurycercus, Leptodora, Cypris, Cyclops,

and Canthocamptus.

To the comparative anatomist, Polyodon is peculiarly not-

able as amoug the oldest of fishes, distinguished, when com-

pared with higher species, by the persistence of juvenile charac-

ters ; and similarly we find that the most remarkable feature

of its food is one which it shares with young fishes in general.

This is, however, a simulated correlation, the food habit not

being due to a persistence of youthful structures of alimen-

tation, but to a remarkable specialization of the apparatus of

food prehension. It must consequently be correlated with a

superabundant supply of minute animal life when and where

these structures originated, or, at least, when they took their

present form ; and taken together with the great size of this

fish and its out-worn dental furniture, seemingly indicates a rad-

ical change in the feeding habits of the species, and a capacity

for adaptation to new circumstances which possibly accounts

for its long survival.



468 Illinois State Laboratory of Natural History.

FOOD OF SMALLER FAMILIES.*

[n'umber of

Specimens Examined. 10



The Food of Fresh-Water Fishes. 469

Food of Smaller Families.—Continued.

Number of

Specimens Examined 10

1



470 Illinois State Laboratory of Natural History.

FOOD OF CATOSTOMATID^.



The Food of Fresh-Water Fishes. 471

Food of Catostomatida!:.—Continued.

N'UMBER OF

ECiMEKS Examined,



472 Illinois State Laboratory of Natural History.

FOOD OF SILURIDiE.



Tlte Food of Fresh-Water FiftJies. 473

Food of Silvhidje.—Continued.



EKEATA.^^

Page 5, line 3 of table, second column, for 39 read 38 ; line 6, second
column for 121 read 120.

Page U, line 17. for conjunction read conjuyation.

Page 21, line 13, for Ricciaci<v read Ricciacea'.

Page 67, line 17 from bottom, for fraligifolia read fraqilifolia.

Page 123, line 4 from bottom, and page 12(5, line 1, for Tricholeoi read
TrichoGolea\

Page 126, line 2, for Tricholea read Trichocolea.

Page 177, line 16, for Lecythia read Lecythea.

Page 333, line 1, after Tachidius add Lilljeb.

Page 33S, under Daphnella brachyura, line 16, insert ffa6.—Massa-
chusetts (Birge), Minnesota (Herrick).

Page 340, line 5, for Hcapaoleberis read Seapholeberis.

Page 389, line 7 from bottom, for carjiogonium read sporocarp ; lines

9, 12, 15, for ojgonium read carpogonmm.
Page 391, line 1, for Cessatii read Cesatii.

Page 400, line 4, for Myceliumin conspicuous read Mycelium incon-

spicuous ; line 14, for coleosporium read Coleosporitun.

Page 401, line 9, for connatas read connnta; line 12, for Taraxicum
read Taraxacum.

Page 408, line 15, for macrocarpa read macrospora; line 18, for Hy-
pohyllous read Hypophyllous.

Pages 470 and 471, head of column 11, for cyprinella read cypri-

nellus.

Page 503, lines 8, 14, and 17, for cyprinella read cyprinellus.

For additional errata see page 247.