-ol

'M MISSION

FORBES .S.A.

Report on Aquatic Invertebrate
Fauna of Yellowstone Park and
Adjoining Regions.. 1893

EXTREME CARL

mis volume is damaged or brittl
and CANNOTbe repaired!

• photocopy onlyifnecesi

• return to staff

do not put in bookdrop

GersteifT Science Information Centre

m

OMMISSION

A PRELIMINARY REPORT

ON

Aquatic Invertebrate Fauna of the Yellowstone

National Park, Wyoming, and of the

Flatliead Region of Montana.

S. A. FORBES,

I'KOH H ILLINOIS.

.1 |>iil,lir:ilim,. April 'J'.l

\SHTNGTO

no v i

1 s

BY
F

[ARTICLE 6.— EXTRACTED FROM THE BULLETIN OF THE U. S. fetSH COMMISSION
FOR 1891. Pages 207 to 258. Plates XXXVII to XLII.]

A PRELIMINARY REPORT

ON THE

Aquatic Invertebrate Fauna of the Yellowstone

National Park, Wyoming, and of the

Flathead Region of Montana.

BY

S. A. FORBES,

PROFESSOR OF ZOOLOGY, UNIVERSITY OF ILLINOIS.

[Date of publication, April 29, 1893.]

WASHINGTON :

GOVERNMENT PRINTING OFFICE.
1893.

S AVICES

6 -A PRELIMINARY REPORT ON THE AQUATIC INVERTEBRATE FAUNA
OF THE YELLOWSTONE NATIONAL PARK, WYOMING, AND OF THE
FLATHEAD REGION OF MONTANA.

BY s. A. FORBES,

Professor of Zoology, University of Illinois,

INTRODUCTORY.

The immediate impulse to the investigation of the aquatic invertebrate fauna of
Wyoming and Montana, here reported in a preliminary way, was supplied by the
ichthyological work of Dr. David S. Jordan, in the Yellowstone National Park, in
1889, and of Prof. B. W. Evermann, in Montana and Wyoming, in 1891.

The waters of Yellowstone Park had been reconnoitered by Dr. Jordan for the
special purpose of ascertaining precisely which of them were destitute of fish and
what was the cause of their barrenness. This having proved to be topographical in
every case — some physical barrier to the entrance of fishes from below — it seemed
possible to stock these waters permanently with valuable game-fishes, and thus greatly
to increase the attractiveness of the Park to a considerable class of travelers. Pre-
liminary to this, however, it was evidently desirable that a full knowledge should be
had of the variety and abundance of the lower animal life of these fishless waters,
since upon this the fishes introduced must chiefly depend for food. To this practical
end it was the wish of Hon. Marshall McDonald, United States Commissioner of
Fish and Fisheries, that my own investigations made in 1890 should be immediately
directed ; but with the understanding that the opportunity thus afforded for a general
zoological survey of the waters of Yellowstone Park should be improved to the best of
my ability.

My associate in 1890 was Prof. Edwin Linton, of Washington and Jefferson College,
Pennsylvania, who, although specially charged with another duty, that of a study of
the parasites of fishes in these waters, rendered me constant and invaluable service in
my own special field.

In 1891 it was niy general purpose to cooperate with Prof. Evermann in an
exploration of the waters of Montana and Wyoming, to be made with reference to
the location of a fish-hatchery; but in this, as in the preceding year, I made every
effort to become as thoroughly acquainted with the animal life of the waters which
I examined as the brief time spent in each locality would permit.

207

208 BULLETIN OF THE UNITED STATES FISH' COMMISSION.

TRIP OF 1890.

Leaving the University of Illinois July 11, I was joined in Chicago by Prof.
Linton July 14, having spent the interval in supplying deficiencies in our outfit. We
left Chicago on the evening of the 14th, reached Mammoth Hot Springs, in Yellow-
stone National Park, via the Northern Pacific Railroad from St. Paul, during the
afternoon of July 17, and went into camp the same evening on Swan Lake Plateau,
with everything ready for the field. Our party at starting consisted of Prof.
Linton and myself, our guide, Mr. Elwood Hofer, who had made our camp ready in
advance, and a teamster, two packers, and a cook. Our outfit contained (besides the
necessary camp equipage) pack animals and saddle horses for six men, a portable can-
vas boat with two pairs of oars, two naturalist's dredges with rope, a set of portable
sieves for assorting the contents of the dredges, a sounding line, a very deep trammel
net 50 yards long, a creek seine, an ordinary minnow seine, a Baird collecting seine,
surface nets, hand nets, two deep-sea thermometers, a dissecting microscope, a com-
pound microscope with complete equipment for field microscopy and for the preserva-
tion of perishable minute material, tanks of alcohol, bottles, vials, etc.

Breaking camp on the morning of July 18, we rode 25 miles through Norris Geyser
Basin and down the Gibbon River to the branch of the latter known as Canyon Creek,
where we camped for the night and made our first collections with hand nets from that
stream. On the 19th we rode through the lower and upper geyser basins and camped
just beyond the latter, on the banks of the Firehole River.

Collections with hand and surface nets were made here from various points on the
Firehole and from the outlet of a warm spring on its banks. As we were now to travel
for some weeks by mountain trails, the teamster was here turned back, and the pack
animals were loaded for the trip across the " continental divide." Leaving this camp
on July 20, we crossed the divide through Norris Pass and went into camp on the
shore of the north end of Shoshone Lake, at the mouth of Heron Creek. A hurried
dip with surface nets was made, in passing, into the waters of some large ponds, with-
out outlet, in the mountains near the summit of the divide.

On Shoshoue Lake we stayed for the three days following (two of the party
circumnavigating it on the 22d), and made extensive collections along shore, in the
inlet of the lake, in an overflow lagoon or pond beside it, and from its own waters with
towing net and dredge, from the surface by day and night, and from the bottom at
depths varying from 8 to 40 feet. Breaking camp on this lovely lake, which will ever
have a peculiar charm in our memories as the place where systematic work on the
invertebrate life of the waters of the Park began, we went on the morning of the 24th
to Lewis Lake, 12 miles below, two of the party running the rapids of Lewis River in
the boat. We camped on the east shore of Lewis Lake, working July 24 and 25 with >
dredge and small nets in the lake, and making miscellaneous collections from streams
of various temperatures and from the waters of a swamp which becomes connected
with the lake in spring.

From Lewis Lake we rode to Heart Lake, a distance of 7 or 8 miles along the
foot of the Red Mountains. Arriving at noon of tbe 26th, we crossed Witch Creek
and camped in a grove of pines above its mouth, not far from the foot of Mount Sheri-
dan, whose precipitous front was a maze of roaring streams supplied by the melting

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 209

snows on its upper slopes. The situation here is one of the most attractive in the
Park. Camping ground and feed are good, water is abundant and excellent, fish of
three kinds — trout, chub, and sucker — are plenty in the lake, and minnows can be
taken by the half bushel in the warm waters of Witch Creek. The place is absolutely
retired (there was not even a trail by the way we came) and quite off any line of travel
even proposed by the Park authorities. The lake is a gem of beauty, a fit companion
to the noble mountain, from whose heights a view of lakes and rivers and mountain
peaks and ranges may be had second to none in this part of the Rocky Mountains.

Our stay in this charming spot extended to nearly five days, all of which but one
were spent in continuous collecting from the lake and from the tributary already men-
tioned as Witch Creek. Violent winds made it difficult to work far from shore in our
light canvas boat, but, with the aid of a small raft made for the occasion, we got good
soundings and dredged successfully about a quarter of a mile out. Here, besides the
kinds of collecting already specified, we used our small seine in Witch Creek and the
trammel net in the lake, taking in the latter considerable quantities of all the larger
kinds of fish the lake supports, in places where the rocky bottom would have made
seining impossible even with a much more cumbrous apparatus.

We shifted camp on July 31 to the west bay of Yellowstone Lake, passing Rid-
dle Lake on our way and pitching our tents on the shore, a few rods above the Upper
Geyser Basin of this bay. Here a line of soundings was run out about 2,000 feet,
the dredge was hauled from the boat 1,000 feet from shore at a depth of 102 feet,
with a bottom temperature of 46°, and at various lesser depths near shore. Other
collections were made from the lake in the usual variety, and also from several of
the warm springs and their outlets. The first water birds were shot here for a
study (by Prof. Linton) of the relations of the fish-eating birds to the parasitism of
the trout, and descriptions were made of rotifers and protozoa which it was not
possible to preserve for later study. A short excursion from this camp gave us
access with the boat and our lighter apparatus to Duck Lake, a land-locked body of
water, formerly connected with Yellowstone Lake, but having now neither inlet nor
outlet at any season of the year. At the foot of Yellpwstone Lake, where we arrived
August 5, our party was reorganized by the dismissal of the guide and pack train and
the engagement of a teamster and saddle horses for the remainder of the trip.

From this point we worked on the lower lake, on Yellowstone River at the outlet,
and on Pelican Creek and smaller tributaries, until the 12th of August, Prof. Linton
going for pelican to the head of the lake, in a skiff, on the 9th and 10th, with a.volun-
teer party from the lake hotel. Towing-net collections were made by this party not
far from the inlet. The dredge was run from a skiff off the landing, on this visit, at
a depth of 100 feet, and also in shallower water. Being unable to reach deeper
water for want of a line left at Norris Geyser Basin, and needing also other supplies
left there, for which I was unable to get transportation to the lake, we left the lake
for a time, starting to Norris Basin on the 12th. The 13th was spent at the Trout
Creek camp, collecting in waters of various temperatures from Alum Creek, above
and below the remarkable hot-spring basin through which this stream flows.

The occurrence of small trout in the upper course of this little creek seemed at

first a mystery, since they are found above the hot springs which boil up in its bed

for a distance of several rods, and so make its waters there altogether intolerable to

fish; but it finally appeared that when the streams are filled by melting snow in

F. C. 13. 1891 14

210 BULLETIN OF THE UNITED STATES FISH COMMISSION.

spring aiid early summer the temperature is so reduced, even in its hottest part, as
to make it possible for fish to pass.

On the 14th we unloaded our boat for a few hours' work on Mary Lake, a clear
and pretty sheet of water lying near the summit of Mary Hill, at an elevation of
8,200 feet, and we thoroughly examined also the upper waters of Nez Perc6 Creek
as we passed down the stream to the Lower Geyser Basin. The 15th we spent in
collecting from the lower part of Nez Perce Creek and from the Magpie, its principal
affluent, for a short distance above its month. Every condition was found suitable in
these streams for the maintenance of fish, and a report to the Commissioner from the
field to this effect was followed by a consignment of Von Behr trout, set free in the
Nez Perce by Mr. Lucas, of the Commission.

August 16 was spent on the Firehole Eiver, from the mouth of the Nez Perce" to
the junction of the Firehole and the Gibbon, and the 18th on the same stream above
the Lower Geyser Basin, our collections ranging from the ford above Old Faithful to
the Middle Geyser Basin below the Excelsior Geyser. Excellent opportunities were
given here for a study of the effects of the geyser outflow on the animal life of the
stream. Some hauls along shore with the surface net were made, in passing, from
Goose Lake, near the lower basin.

On the 17th we collected at Canon Creek again, and from the Gibbon Eiver, at
the mouth of this creek, and also above and below the falls of the Gibbon, well
known as an impassable barrier to the movement of fishes up this stream. In our
collections above and below falls in these rivers and creeks, it was my object to learn
whether any other animal inhabitants of these streams were similarly excluded from
their upper waters.

On the 20th we made a trip from Norris Geyser Basin to the Twin Lakes (in one of
which whitefish had already been planted by a Fish Commission party), returning by
a small lakelet without outlet, called the Lake of the Woods. We went thence to the
Grand Canon, collecting by the way from the Gibbon above Virginia Cascade, and
from minor waters passed, and on the 22d made a trip to a lake nameless to the guides,
but marked Grebe Lake by the geologists, and mapped as the source of the Gibbon.
We carried boat, dredge, small seines, and our lighter collecting apparatus to this
lake, and thoroughly overhauled it, as typical of its kind.

Returning to Yellowstone Lake on the 23d, we explored Pelican Creek on the 24th
for several miles above its mouth, and on the 25th dredged from two rowboats at a
depth of 195 feet, with a bottom temperature of 42.5°. Returning next day to the
canon, we collected from the lower course of Alum Creek, the upper part having been
explored by us previously. On the 27th, sending the wagon to Mammoth Hot Springs
by the traveled road, Prof. Lin ton and I took the trail down the river to Yancey's
Ranch, crossing Mount Washburn, and making considerable collections from Tower
Creek above the fall. On the 28th we went from Yancey's to the springs, stopping
by the way at Lava Creek and Blacktail Deer Creek. On the 29th the usual col-
lections were made from the Madison, at the crossing of the Cooke City road, and
also from Swan Lake on the plateau of the same name, to which our boat and the
usual collecting equipment were transported for us by Capt. Boutelle, U. S. A., acting
superintendent of the Park.

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 211

On August 30 we closed the field work of this trip with collections from Glen
Creek below the falls, from the Gardiner River at the mouth of Hot Creek, near Mam-
moth Hot Springs, and from a small lakelet among the hills, towards Gardiner.

The accumulations of the trip were made under 387 collection numbers, represent-
ing 43 localities.

Our work was limited substantially to the central park plateau, only that about
Mammoth Hot Springs passing beyond the lava formations which cover the plateau
everywhere to an unknown depth and noticeably affect, as we discovered, the animal
life of its waters. The river systems investigated were those of the Gardiner, the
Madison, and the Yellowstone, on the Atlantic side of the "continental divide," and
of the Snake on the Pacific slope. The principal fishless waters examined were Sho-
shoneand Lewis lakes, the Upper Gibbon and connected waters, the Firehole and its
branches, Goose Lake, Twin Lakes, Swan Lake, and Tower Creek. The effects of
geyser and hot-spring outflow were shown especially by collections made from the
Firehole and from Alum Creek; and those of the occurrence of falls in the course of
these mountain streams were shown especially by collections from the Gibbon and
some of its tributaries. The highest elevation represented by our aquatic material
was that of Mary Lake (8,200 feet) and that of a small lakelet near Norris Pass,
not far from the same level. The greatest depth at which we dredged was 195 feet
in Yellowstone Lake, although this depth was exceeded somewhat in the work of
the following year. The altitude of this lake is 7,740 feet above the sea.

As material for a study of variations in biological condition, we obtained an
abundance of specimens for a comparison of the system of life in lakes, ponds, rivers,
and creeks where no fish are found with those in which only a single species occurs,
and with those supporting from three to eight kinds of fishes.

The effect of the "continental divide" or watershed upon the distribution of
aquatic animals is, of course, amply illustrated by our material; and this, taken in
connection with materials gathered the following year from lower altitudes, should
show something of the limitation of range of several species imposed by differences of
elevation and the like. The influence of widely different geological conditions should
likewise become manifest as we compare the animals of the waters of the Park
plateau with those outside.

My warmest thanks are due to Capt. F. A. Boutelle, acting superintendent of
the Park, who encouraged and aided our investigations in every possible way, and
to our guide, Mr. Elwood Hofer, whose tireless energy and active personal interest in
our operations were greatly in our favor. He was not only the guide and manager of
our movements, but a most efficient volunteer assistant in camp and in the field.

212 BULLETIN OF THE UNITED STATES FISH COMMISSION.

TRIP OF 1891.

Leaving Champaign, Illinois, accompanied by my university assistant, Mr. H. S.
Erode, on the afternoon of August 10, I arrived at Livingston, Montana, on the
evening of the 13th, by way of Chicago and St. Paul, and proceeded thence to
Helena, to which place our outfit had been shipped from Washington and Champaign.
Delay in the arrival of part of the equipment made field work impracticable until
the afternoon of the 17th, when we made our first collections from the Jocko River
at Eavalli, on the Northern Pacific Railroad in western Montana.

My immediate object on this part of our trip was an investigation of Flathead
Lake and its tributaries, and among these especially a small, very cold trout stream
previously visited by Prof. Evermanu, and noted as suitable for the supply of the
proposed trout hatchery.

Flathead Lake itself offered a very interesting contrast to Yellowstone Lake,
examined the year preceding, but with sufficient resemblance also to make compar-
ison instructive. It was then commonly reached by stage from Ravalli through the
Flathead reservation — a trip which we took on August 18, going thence by steamer
to Demersville on Flathead River, about 25 miles above the lake. We began our
collections from this river on the evening of the same day, and worked here also on
the 19th, collecting especially from bayous and backwaters. Through the kindness
of Mr. H. W. Parchen, president of the Helena Board of Trade, and of his associates
in a game and fishing club, I had the use of a small steam launch for the work on
the lake — an indispensable advantage, without which we could have done only a little
imperfect alongshore work. Accepting the cordial invitation of this club to make
their club house our headquarters, we went thither from Demersville in the launch
August 20, and made our first surface-net collections in the afternoon of the same
day. This club house is built upon a large bay at the upper end of the lake, partially
sheltered from the rather violent winds prevailing, and yet containing water of suffi-
cient depth to illustrate fairly the deep-water conditions of this lake. It represented
also every variety of shore and bottom — sandy flats, weedy shallows, rocky shores,
and gravelly banks — and had the further advantage, for our purpose, of giving ready
access to a considerable tributary of the lake, named Swan River on the map, but
locally known as the "Big Fork." Our collections here continued over the three fol-
lowing days, and included surface-net work of all varieties, many alongshore collec-
tions, and several hauls with the dredge, made by aid of the launch, in water ranging
from 80 to 162 feet in depth. Considerable collections were also made on Swan
River, especially upon the rocky rapids a short distance above its mouth.

On the 24th we made a horseback trip to Swan Lake, 12 miles above our quarters,
and spent several hours collecting with our smaller apparatus from the lower part of
that lake and from a cold trout stream emptying into the river a short distance below.

On tke 25th we went by the regular steamer to the foot of Flathead Lake, where
we made such collections from this shallow southern end of the lake and from its
outlet (the Cosur d'Alene) as a heavy storm would permit, finishing our work in
this region on the 26th, and starting for Helena and Yellowstone Park. Our work in
the Park was confined to the northeastern part — not visited in 1890 — and to Yellow-
stone Lake, to which I went especially for a more thorough use of the dredge (with
the aid of the passenger steamer) than I could make from skiffs the year preceding.

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 213

From Yancey's Ranch, which we reached on the 29th, we explored Slough Creek
above the lower rapids, and some alkaline ponds near Baronette's Bridge, and on the
31st went up the East Fork of the Yellowstone to Soda Butte Station on the creek of
the same name. Collections were made on the way from Amethyst Creek, and from
the East Fork of the Yellowstone, where this creek empties into it. September 1 we
spent near Soda Butte Station, at work in the creek and in Trout Lake 2 miles
north of the " station." Returning to Yancey's September 2, we examined the over-
flow waters of the creek and searched the East Fork thoroughly at the Soda Butte
bridge, and finished our collections from the river in that vicinity.

On the 3d and 4th we continued to the cafion and to the lake. The 5th was
spent in making shore collections from Stevenson Island in Yellowstone Lake
and in an examination of the small ponds and bayous of the island itself. On the
forenoon of the 7th we finished our work on Yellowstone Lake by making three hauls
of the dredge from the little steamer Zillah in the vicinity of Stevenson Island, at
depths varying from 20 to 198 feet.

The work of the season closed, September 10, with collections made from two
localities previously examined by Prof. Evermann with reference to establishing a
fish-hatchery — Bridger Creek and a cold spring adjacent, near Bozeman, and some
springs and small streams near Boteler's Ranch, just north of the Park. The return
trip was made by the Northern Pacific Railroad, September 10 to 13.

The collections of this summer were made under 73 collection numbers, represent-
ing 23 localities.

Apart from the practical points aimed at, and the opportunity to further extend
our knowledge of the aquatic life of Yellowstone Park, a region whose zoology must
long have an exceptional interest, I value the results of this year's work chiefly as
affording the means for a comparison of the animal life of two lakes so similar in
many respects as Flathead and Yellowstone, and yet widely contrasted in altitude, in
geological surroundings, and in topographical and geographical relations. It is, in my
judgment, by a thorough examination and critical comparison of typical situations like
these that the most interesting and immediately fruitful additions to zoological science
are to be made in this field. I have only to wish that a longer stay on each of these
lakes might have made possible a more minute and exhaustive study of their animal
life and its relations to varying conditions of depth, bottom, temperature, season,
weather, bionomic association, and the like.

DISCUSSION OF THE COLLECTIONS.

While the partial and, in most cases, merely preliminary way in which the mate-
rial of these expeditions has as yet been studied makes any full discussion of the
results impossible, it seems best that a report of progress should be made, presenting a
summary review of the invertebrate life of these waters in the midsummer season, with
descriptions or determinations ot such new or particularly abundant and important
kinds as have thus far been made out. Such a statement will include, in fact, the
greater part of the economic results of immediate utility, and may be said, therefore,
to fulfill the leading object of the work. This report may be most conveniently cast
in geographical form, the life of each river system being separately discussed ; but, for
want of time to examine the entire mass of the collections, only a preliminary account
of the fauna of the still waters visited, from temporary pools to Flathead and Yellow
stone lakes, will be given at present.

214 BULLETIN OF THE UNITED STATES FISH COMMISSION.

•

The systems to which the various waters examined belong are those of Snake
River and the Columbia on the Pacific slope, and of the Yellowstone and the upper
Missouri on the Atlantic slope. The first is represented by collections made in the
southwest part of Yellowstone Park, the second by those from the Flathead region, the
third from the north and eastern parts of the Park and from the vicinity of Boteler's
Ranch, and the fourth by those from the branches of the Madison in the central-
western part, and from Bridger Creek near Bozeman.

The collections now reported were made from the waters named in the following
list : A mountain pond near Norris Pass, Shoshone Lake, Lewis Lake, Heart Lake,
Yellowstone Lake and certain of its tributary waters, Duck Lake (near the Yellow-
stone), Mary Lake, Goose Lake, Twin Lakes, Lake of the Woods, Grebe Lake (at
the head of Gibbon River), Swan Lake (Yellowstone Park), a lakelet near Mammoth
Hot Springs, Trout Lake (near Soda Butte), small ponds in the Soda Butte bottoms,
alkaline ponds near Baronette's bridge and several other scattered ponds, Flathead
Lake, and Swan Lake (Montana).

THE SNAKE RIVER SYSTEM.

This system was reached only in its head waters, all a few miles from the low "con-
tinental divide," which sometimes separates only imperfectly the waters of the Pacific
side of the continent from those of the Atlantic slope. Shoshone and Lewis lakes of
this system are, respectively, about 1 j and 2 miles in a direct line beyond the crest of
the divide, and Heart Lake is less than 4. From these lakes and from their tribu-
taries all the collections made in this district were obtained, with the exception of a
little group snatched with the hand net, while the pack train was passing, from a
mountain pond near Norris Pass, on the Shoshoue trail.

This pond was completely stagnant and filled with growing vegetation (including
filamentous and gelatinous algae) and fallen timber. The collection contains little to
indicate the elevation from which it came, but is of interest in comparison with the
contents of the very different waters of Shoshone Lake a few miles away and 400
feet below. In this pond I found a small spotted larval salamander, with both fore
and hind legs already budded out, a considerable number of young insects (Corisa,
Chironomus, and Gorethra — larva and pupa), an amphipod crustacean (Allorchestes
dentata), and a great number of entomostraca. Among the latter were Diaptomus
shoshone,* two species of Cyclops, Daphnia pulex, an undetermined species of Daphnia,
and a Geriodaphnia. A black springtail (Poduridce) and a wheel animalcule (Lacinu-
laria socialis), occurring abundantly in globular colonies, were the only other animals
recognized in this preliminary examination.

Shoshone Lake. — Shoshone Lake is a lovely little body of clear blue water lying at
the level of 7,740 feet — almost exactly that of Yellowstone Lake. It is shaped like a
blotted T, with the stem, 7 miles long, extending north of east from the Geyser Basin,
at the head of the lake, and the crosspiece, at the eastern end, about 4 miles in length.
The stem reaches a width of 1 J nules, but narrows near the middle of the length of the
lake to less than half a mile. This lake lies charmingly secluded in a hollow of the
densely wooded mountains which surround it everywhere except to the southeast. It
is at present accessible only by mountain trail from the Upper Geyser Basin, and has
fortunately been omitted from the system of improved roads now being made for wagon

* Described on pagu 251.

x AQUATIC INVERTEBRATE FAUNA OF* WYOMING AND MONTANA. 215

travel. The shores are bold but not much broken, steepest on the south and west,
where the 8,000- foot line runs a quarter to half a mile from the margin of the lake.
On the northeast a peak about half a mile back rises to 8,GOO feet, and others nearly
as high lie not far north of the eastern end. To the south the Pitchstone Plateau lifts
its black and forbidding mass — patched with snow all summer — to a height of nearly
9,000 feet; and to the southeast, 7 or 8 miles away, but seemingly less than half as
far, the Eed Mountains rise, culminating in Mount Sheridan, 10,200 feet above the sea.
The rampart of hills surrounding the lake opens out on the northeast, where Heron
Creek comes in ; on the west, to form the valley of Shoshoue Creek ; on the south,
where Moose Creek drains a swampy tract about 1 mile across and 5 miles long; and
on the southeast, where the waters of the lake pass out through Lewis River. Some
smaller tributaries empty at the Geyser Basin, on the eastern end; and a number of
little rivulets, dry at times, drain the hills at various points. Near the mouths of the
larger streams, ponds or small lagoons occur, connected with the lake at high water,
and in midsummer thick with vegetation and swarming with animal life. The imme-
diate shores are commonly rocky except for an occasional narrow beach of black
volcanic gravel. There is little weedy water in this lake, the sandy bottom bearing
at best a sparse growth of Potamogeton and plants of similar habit.

The only soundings made by us were in the nortli arm or bay of the lake, where
depths of 40 and 50 feet were reached from a third to half way across the mouth of this
bay, starting from the eastern side. The bottom at these depths varied from sand to
soft mud, the latter without vegetation, the former with a growth of Cladophora.

Our camp was placed in a small grove on the Hat at the mouth of Heron Creek,
where we had at hand the creek itself and a small, very weedy, and very muddy lagoon,
filled earlier with overflow waters, but then disconnected from the lake. Our collec-
tions were made chiefly in the north bay of the lake, but a few things were taken
from the western end, and a few collected alongshore as we made our way to the
outlet. In the north bay, besides making collections along shore and with hand nets
in the shallow water, we hauled the surface net repeatedly from the boat, from 8 a. m.
to 9 p. m., in both clear and rainy weather, and dredged at various depths from 6 to 40
feet. Our larger apparatus was useless, as there were no fish in -this lake.*

* Here we first heard, while out on the lake in the bright still morning, the mysterious aerial sound
for which this region is noted. It put me in mind of the vibrating clang of a harp lightly and rapidly
touched high up above the tree tops, or the sound of many telegraph wires swinging regularly and
rapidly in the wind, or, more rarely, of faintly-heard voices answering each other overhead. It begins
softly in the remote distance, draws rapidly near with louder and louder throbs of sound, and dies
away in the opposite distance ; or it may seem to wander irregularly about, the whole passage lasting
from a few seconds to half a minute or more. We heard it repeatedly and very distinctly here and at
Yellowstone Lake, most frequently at the latter place. It is usually noticed on still, bright mornings
not long after sunrise, and it is always louder at this time of day ; but I heard it clearly, though
faintly, once at noon when a stiff breeze was blowing. No scientific explanation of this really
bewitching phenomenon has ever been published, although it has been several times referred to
by travelers, who have ventured various crude guesces at its cause, varying from that commonest
'catch-all of the ignorant, "electricity," to the whistling of the wings of ducks and the noise of
the "steamboat geyser." It seems to me to belong to the class of aerial echoes, but even on that
supposition I can not account for the origin of the sound.

216 BULLETIN OF THE UlftTED STATES FISH COMMISSION.

The assemblage of animals in this lake offered a peculiarly interesting subject of
study, since it included practically no aquatic vertebrates. There were, of course, no
fishes at all ; we saw no turtles or water snakes, but a single frog, and only one small
salamander. The dominant groups were insect larvae, leeches, amphipod crustaceans,
and entomostraca. By far the most ab undant aquatic insects were caseworms (larval
Phryganeidw), mostly pupae at the time of our visit. There were no crayfishes and
no isopod or phyllopod crustaceans; but two amphipod genera (Oammarus and
Allorchestes) were very abundant. The Gammari, represented by a single large and
robust species (Oammarus robustus Sin.), were exceedingly common, creeping or swim-
ming about, on or near the bottom, inshore, especially where collections of debris from
the inlet rested in hollows of the rippled sand. They sometimes rose to the surface
at night, where our towing net occasionally took a surprising number of them — at
one haul almost nothing else. They occurred also abundantly i n our deepest dredg-
ings, in the lagoons examined and in the streams flowing into the lake.

This lake seemed, indeed, a paradise for the Gammari, containing an abundance
of food for them, both animal and vegetable, fresh and in process of decay, and
scarcely anything that fed upon them in turn. The commonest large leech (Nephelis
obscura, var. maculata) feeds upon them to some extent, as I found by the dissection
of two specimens; but another of these leeches voided a large horse-fly larva
(Tabanus). Their own food, if I may judge from that of seven specimens which I
dissected, is quite varied, consisting of rotting vegetation (whose condition was shown
by the mycelial threads running through it), of fresh algiB, and other green-plant
substance, and of entomostraca (Diaptomi as far as seen). The stomachs of three con-
tained, also, a noticeable amount of pollen grains of the pine. In three of the seven
specimens examined large numbers of Grcgarlncc infested the intestine. Their prob-
able effect was shown by the fact that the intestine was empty in two of the para-
sitized specimens. Our Gammarus was thus practically at the head of the biological
system of this lake, which was for it a royal domain where it was free to exact tribute of
all, yielding scarcely anything itself in turn. The females at this time had their brood
cavities loaded with young.

The entomostraca were principally a single species of copepod — a very large blood-
red Diaptomus, new to science and here described as D. shoshotie. This occurred in
great numbers in several hauls with the surface net, and could usually be seen on a
calm evening near the surface, where its tiny sportive leaps in the air kept the water
microscopically agitated, as if by minute fish. Another Diaptomus, near D. sicilis and
perhaps a variety of that species, occurred much more sparingly, and a third species
of this genus, described on p. 252 as D. lintoni, was less frequently seen than either of
the foregoing. There were a few species of Cyclops here, G. serrulatus Koch, C. gyrinus
Forbes, C. minnilus (new), and perhaps others; also, a Oypris, a Bosmina, a Chydorus,
a Daphnia (D. pulex), and Polyphemus pediculus. So far as the Crustacea were con-
cerned, the lake was in practical possession of Gammarus robustus and Diaptomus sho-
shone.

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 217

The food of the latter little species was peculiar at the time of my visit, and the
collections consequently give little idea of its usual function in the biological system
of the lake. All of more than fifty specimens examined from several of the Shoshone
Lake collections had fed freely, and often greedily, on the pollen grains of the pine.
Only a single specimen dissected contained also some fragments of another entomos-
tracau, among which were single antenual segments of a copepod, probably of its own
species.

This fact is but an illustration of the dependence of the animals of a lake on the
contributions made to its stores of food by the surrounding land. As a surprising-
number of fishes profit largely by the terrestial insects falling into the water, so this
little copepod horde must live for some weeks, to a very large extent, on the pollen of
the surrounding forest, relaxing the pressure, for a time, on the plant and animal life
of the waters, which is doubtless their more usual food resource. Would the destruc-
tion of the forests here seriously diminish the number of Diaptomi, and thus lessen the
food supply of the young of the trout with which this lake has lately been stocked?

The large leeches taken here occur throughout the Park in suitable situations,
and have been noticed by earlier collectors at Yellowstone Lake. At the Shoshone sta-
tion they were frequently seen in the clear shallow water, either swimming actively or
creeping along the bottom. They are carnivorous leeches, as already mentioned,
almost the only native enemies of the Gammarus worthy of notice in Shoshoue Lake.

A frequent and interesting occurrence during our visit was the appearance at the
sunny surface of the lake of a large dark-gray caddis fly (Neuronia sp.) freshly escaped
from its pupal prison and flitting rapidly along with its imperfectly expanded wings,
just on top of the water, going with speed directly for the shore. The number of these
insects — caseworms in the larval state — was shown by the thousands of their empty
cylindrical cases washed ashore. Larva;, pupa1, and imagos were all common at the
time of our stay. The case of this species is composed of thin, irregular pieces of
vegetation .(largely fragments of leaves and epidermis of water plants), or of chitinous
plates of insects, eked out by filamentous alga; and other miscellaneous objects, all
cemented and imbedded in the tough secretion from the salivary glands of the insect
itself. On preparing to pupate, the larva closes the mouth of its tube by a coarse
latticework of hardened mucus which protects the insect within, permitting at the
same time the free access of water. Shoshone Lake, it need not be said, was an ideal
place for the breeding of these caddis flies, since it contained no common carnivorous
animal large enough to attack them.

Chironomus larva? were common in these waters, and their pupae, ready to emerge,
appeared often in the surface net.

The mollusks taken were limited to a few specimens of a large, dark Physa, with
an exceedingly thin and brittle shell, and a small, heavy Pisidium, with a few con-
spicuous lines of growth. These last were mostly empty shells, collected from the
hollows of rippled sandy bottom, where they were readily seen as one floated over in
a boat. An occasional dragon-fly larva (Libellulidw), large larv® of Hydrophilus,
a very few hydrachnids, some slender annelids — undetermined as yet, and very diffi-
cult of preservation — and a considerable collection of digitate fragments of Spongilla
are worthy of mention. So also is the scarcity of waterbugs, limited, indeed, in our
collections from the open lake, to a single specimen each of Notonecta and Corisa.

218 BULLETIN OF THE UNITED STATES FISH COMMISSION.

The small brown lagoon or pond already mentioned as occurring near the lake
was an example of a kind quite common along the lake borders of this region. It is
separated from the lake itself only by a narrow strip of beach, and is largely filled
with pond lilies (Nuphar], grass, algiB, and the like, which grow out of a deep, soft
ooze. There was little in the assemblage of animal forms of this place to suggest its
elevation of more than 7,000 feet, unless the scarcity of mollusks and the higher Crus-
tacea be so explained. A species of Physa and one of Pisidiu.m were the only mollusks
taken. Insects were represented chiefly by ephemerid larvae, larva; of Culex, Chirono-
mus, and other small Diptera, caseworms, Notonecta, Gorisa, Agrion larvae, and larvse
of Dytiscidw; Hydrachnidce by a single species of scarlet water-spider; AmpMpoda
by the lake species Gammarus robustus and Allorchestes dentata; eutoinostraca most
abundantly by a Daphnia of pale pink color, not seen by us before, and here described
as D. angulifera, by D. pulex in moderate numbers, by Polyphemus, Scapholeberis
mucronatm, Eurycercus lamellatus, Chydorus, Cypris, Cyclops gyrinws, G. serrwlatus,
etc., and by no Diaptomi, so far as observed. Leeches were present, although not
numerous — the species already mentioned (Nephelix maculata) and one not detected in
the lake, Aulostoma lacustris Leidy. This pond thus differed from the lake in the
larger number and variety of insects, especially in the larval state, by the absence
of Diaptomi, and by the vast predominance of the new Daphnia. The latter had
evidently been very much more abundant earlier in the season, as shown by the quan-
tities of its summer eggs. These formed a film over many square feet of the surface
and had been washed ashore in quantity as a scum-like deposit along the bank. A
few of the females were still bearing their ephippia.

The collections made by Prof. Linton from the lagoon at the western end of the
lake are similar, as far as they go, but contain no entomostraca.

Lewis Lake. — Lewis Lake is so closely associated with Shoshone that the two
might very well be treated by the biologist as one. The water, shores, bottom, and
surrounding country have substantially the same characters for both, and their free
connection by a river without falls and only some 3 miles in length tends to oblit-
erate any small local differences. The fact that fishes are excluded from both lakes
by falls in their common outlet still further assimilates them in biological condition,
the only noticeable differences remaining being those of size and depth.

Lewis Lake is but 3 miles long by 2J in breadth, with a greatest depth, in our
soundings, of 80 feet. It is rudely triangular in form — more distinctly heart-shaped,
in fact, than Heart Lake itself. Its level, 7,720 feet, is but 20 feet below that of
Shoshone Lake. Its western banks are highest and boldest, the 8,000-foot contour
running usually from a quarter to half a mile from the shore. On the north and
northeast the country is relatively low toward Shoshone and Yellowstone lakes, but
the immediate banks are occasionally bluffy and the shores are everywhere wooded.
The Bed Mountains are close at hand, a range of ten peaks to the southeast; and to
the south looms the great Teton group, the noblest mountain view to be had from
any part of the Park.

There is a small hot-spring basin at the northwest angle of Lewis Lake, and a
swampy tract about half a mile square lies beside it to the northwest, connected with
it for a fortnight or so during high water. At other times communication is prevented
by a narrow strip of beach, sand and gravel, a few feet across. At a little trouble
and expense a permanent passage-way for fishes might be made and maintained, giving
free access to considerable breeding-grounds and stores of food.

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 219

On the west side a small permanent creek came in, about 30 feet wide and 2 feet
deep at the time of our visit, with several acres of somewhat swampy ground at its
mouth ; and here also were three small warm creeks (150° F. ) and one of cold water
(50°), the latter apparently supplied by melting snow from the borders of the Pitch-
stone Plateau. A series of small lagoons, filled with sedge and bulrush, open into the
lake along this shore. The cold waters from the Eed Mountain range are diverted
from this lake by the course of Aster Creek, which drains the northern slope and
empties into Lewis Kiver, about 2 miles below the lake.

Treacherous and stormy weather during our brief stay prevented our making
many collections in the open water. The dredge was hauled, in fact, but once, and
then at a depth of 56 feet, about a quarter of a mile' above the outlet; and even the
surface net was only twice used far enough from shore to give us the so-called pelagic
entomostraca. The remaining collections were gathered inshore, in the swamp
adjoining, and from streams, both warm and cool, on the western side.

In the dredge, at 50 feet, with a bottom of fine black mud and dead vegetable
debris, we took quantities of the large Oammarus, a few Allorchestes, many Chirono-
mus larvae, specimens of Pisidium, and an undetermined annelid not preserved in
condition for identification. The preponderance of Gammarus was as noticeable here
as in the companion lake, although fewer were seen along shore in shallow water. The
same may be said concerning the Diaptomi taken in open water in the surface net.
The gigantic Diaptomus shoshone was relatively much less numerous, however, than
the much smaller ~D. sicilis var. It was feeding extensively upon pine pollen here, as
in the other lake. Although properly pelagic forms — those most at home, that is, in
the open water and in the deeper parts of the lake and found rarely, if at all, in the
small lagoons — these Diaptomi nevertheless extend their range close inshore, where
they might be seen with the naked eye in the water or taken in the net, even when
the surface was decidedly rough. These collections contained many examples of a
peculiar entomostracan (Holopedium gibberum) not noticed in Shoshoue Lake, and also
an abundance of a colonial rotifer belonging to the genus Conochilus — allied to C. vol-
vox, but apparently undescribed. *

In the Irfke, near the entrance of the warm stream on the western side, were quanti-
ties of young water-bugs (Notonecta), an occasional Corisa, Gammarus, Clrironomus, and
ephemerid larvae. Caseworms were also abundant in the lake, and the air alongshore
was full of two species, one black and the other pale brown, just emerged and pairing.

From a small lagoon filled with the overflow waters of a geyser, at the western side,
we took with the hand net several sp( cimens of Gammarus and hydrophilid larva}, a
few Allorchestes dentata and young Corisa, many libellulid larva;, large and small, and
larva} of Agrionincc, a single small Physa, and several water-beetles (Ccelambiis).

From a warm stream at the same place (150° F.) we have many ephemerid larva},
several caseworms and a single hydrachnid, a great quantity of large and vigorous
specimens of Gammarus robustus, and a smaller number of 'Allorchestes dentata, many
examples of Pisidium, a Physa, a few annelids (Oligochcvta), and a single leech
(Nephelis 4-striata) not taken by us before. The Gammarus was feeding very freely
on dead and decaying vegetation and filamentous alga}, with some fresh vegetable
fragments and a little pine pollen. No traces of animal food occurred in two speci-
mens dissected, one half grown and the other of the largest size.

* Described on page 256 as C. leptopus, n. s.

220 BULLETIN OF THE UNITED STATES FISH COMMISSION.

From the swampy tract on the eastern side of the lake already mentioned col-
lections were made by hauling a surface net in the open water, by searching dead
leaves, and by washing off the lily pads (Nuphar] in the net. These waters were
swarming with life, chiefly insect larvae and crustaceans. They apparently contained
relatively few inollusks, several specimens of a large Phym, a few Pisidium and one
Amnicola occurring in our collections. The insect larvie included Agrion, Ephemera,
Chironomus, Corisa, Hydrophilus, and Corethra. The only amphipod crustacean was
Allorchestes dentata, represented by but few examples, but the open water contained
a great quantity of Daplinia angulifera and an occasional Sida crystallina. Among
the lily pads the same Daplinia occurred, together with a great number of Sida, a few
examples of Cyclops and Diaptomi, and several of Vaphnella and of Polyphemus pedic-
ulus (young and adult), and a single short, dark cyprid. Leeches and their capsules
were frequent, the usual spotted and striate species (Nephelis maculata and N. 4-striata).
A single specimen of Clepsine eleyans and another Clepsine not determined were also
noticed. The capsules of these leeches were common on the leaves of water-plants.
Among less conspicuous objects, small Hydrachnidw, Hydra fusca, and the colonial
rotifer Conochilus were abundant.

If we may pause now to glance at the animal life of these three lakes, character-
istic as they are for their region, as compared with that of similar lakes of much lower
altitude — Lake Geneva, in Wisconsin, for example — we find that the large and con-
spicuous differences, so far as invertebrates are concerned, lie mostly in mollusks and
crustaceans. The complete absence of Unionidw, of Paludinidce, Melaniidce, and of
Valvata, and the scarcity of Planorbis and Amnicola are cases in point. The absence
of crayfishes, of Epischura, and of Simocephalus is the most notable distinction in
the crustaceaii list. Polyzoa also were extremely few.

Heart Lake. — Heart Lake had to us the very especial interest that it gave an
opportunity, hitherto unparalleled in this country, to study by comparison the effect
of the presence of fishes on the bionomic system of a mountain lake ; and as the barren
waters of Lewis and Shoshone lakes have since been stocked with trout by the U. S.
Fish Commission, the results of this comparison of native conditions may hereafter be
checked and supplemented by a study of the later state of invertebrate life in these
two lakes. This lake is situated similarly to Lewis and Shoshone, is of nearly the
same size as Lewis Lake, and is in most respects a companion to that and Shoshone,
but differs totally in the fact that its outlet is unobstructed by falls and that it is
consequently well supplied with fishes. It lies only 5£ miles from Lewis Lake, in a
straight line, and about 6 miles from the southern arm of Yellowstone Lake, but the
latter is on the opposite side of the divide and is consequently connected with a
different system of waters. It is divided by a peninsula into two unequal parts, the
larger of which, rudely rhomboidal in shape, is approximately 1£ by 2 miles in diameter.
The smaller part is subtriangular, with principal diameters of about a mile, and the
narrow neck uniting these two divisions of the lake is about a quarter of a mile across.

Heart Lake differs from Lewis and Shoshone by its closer proximity to the Red
Mountains, especially to Mount Sheridan, and consequently by the much greater
amount of snow water which it receives. At the time of our visit, during the last
days of July, the rush of rivulets down the mountain slope, supplied by the melting
snows, filled the air all day with a noise like that of a train of cars. This lake has
also its hot-spring and geyser basin, but receives through Witch Creek a relatively

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 221

larger body of warm water than either of the others. It drains, according to the
published map of the Geological Survey, a larger basin in proportion to its size and
is bordered on the north by a marshy tract 2£ miles long by nearly a mile wide. Its
surface lies 250 feet below that of Lewis Lake and 270 below the Yellowstone. Its
waters are very clear, but are nevertheless much more weedy alongshore than those
of either of the other lakes.

The slope of Mount Sheridan continues downwards into the lake a little distance,
and the water consequently deepens rapidly from the eastern shore. About 200 feet
out the depth was 94 feet; at 400 feet it was 124; and at 1,000 feet it was 146. The
bottom temperature at this latter depth was 40° F.

Our camp was pitched on the western side, about half a mile from the mouth of
Witch Creek, and our work was confined to this shore and to a distance of about half
a mile along the northern shore. Our dredgings here were made in three localities:
in shallow water inshore, at a depth of about 10 feet; upon rocks a little distance out,
at a depth of 30 feet; and in deep water from 46 to 120 feet, with a bottom of soft mud.
Collections were made with the surface net from the open water at various hours of
the day from 9 a. in. to 9 p. m., under such conditions of weather as offered themselves,
and also from shallow water among weeds, commonly near the bottom. In addition
to these, considerable collections of fishes were made with the trammel net and the
smaller seines, the latter of which we used in Witch Creek as well as alongshore in
the lake itself, and from these fishes a quantity of material was obtained for a study of
the food of the various kinds.

As might be supposed, some noticeable differences appear on a comparison of our
collection lists, some readily accounted for and others at present inexplicable unless
as the secondary or more remote effects of the first. It is naturally to be expected
that in so small a lake, and one with so few opportunities for successful concealment
or escape, the kinds of invertebrates on which fishes feed by preference would be
unable to maintain themselves in as large numbers as in similar situations where
fishes do not occur at all; and especially will this necessarily be true if we find that
the fishes destroying these invertebrates are not strictly dependent on them for food,
but eat other things as well. This is true of both the trout and the sucker, the former
being almost indiscriminately carnivorous, and the latter mixing insect larvae and the
like with a large proportion of vegetable food.

It is probably in this way that we are to explain the fact that we did not find in
our stay on this lake a single larva of Neuronia (the largest caseworm in these waters),
so abundant in Shoshone Lake, nor a single amphipod crustacean (Oammarm or
Allorchestes) — all large enough to afford an attractive food to one or all of the fishes in
these waters. That they occur here I can scarcely doubt, although the distribution
of the Gammarus seems at best very whimsical in this region, but they certainly were
far less common than in the adjacent lakes. The absence of the larger leeches (Neph-
elis maculatus) may be due to our failure to find suitable places for them, or they also
may be eaten by fishes.

More difficult to understand is the very remarkable fact that we did not find here
so much as a single specimen of the almost gigantic copepod, Diaptomus shoshone,
although its companion elsewhere, the smaller species of Diaptomm, was extremely
abundant in all our open-water hauls. Equally difficult of explanation was the vast
abundance of the entomostracan Daphnella brachyura — not once taken before we

222 BULLETIN OF THE UNITED STATES FISH COMMISSION.

reached this lake, and here the most abundant species next to Diaptomus. With it
was Leptodora hyalina, also extraordinarily common (for a predaceous species) in the
product of every haul. The absence of Polyphemus pediculus is less remarkable, since
it was not common in the 'other lakes, and the fact that Holopedium gibberum should
be wanting here while common in Lewis Lake is without special meaning, since we
failed to find it in Shoshone Lake.

A fuller discussion of this matter must be postponed until our materials illustrat-
ing the food of fishes taken here have been completely studied ; and I will merely add
at present some further details concerning the general collections.

The beach on the west shore was gravelly, in some places with boulders of consid-
erable size, and occasionally with a stretch of sand. The bottom was covered with a
growth of weeds (Potamogeton, algae, etc.) and was greatly cumbered with driftwood.
Our deepest dredging was made oft' this shore, beginning at a depth of 120 feet, with
a temperature of 40° F., and ending at 46 feet, with a temperature of 53°. This haul,
made from a raft and boat together, was about 100 yards in length, over soft mud
containing some very tine sand, but consisting largely of organic debris, both vegetable
and animal. The latter, minutely examined, was seen to be made up of the valves of
entomostraca, fragments of the cuticle of insect larvse, and the shells of rhizopods
(Difflugia and Echinopyxis), while the vegetable remains consisted of minute pieces of
Vaucheria and other filamentous alga), and fragments of higher plants, with a multi-
tude of shells of diatoms. The living animals of this haul were Chironomus Iarva3 and
pupae only, the former red, and many of them in their usual tubes composed of mud
and minute organic remains.

The shallower hauls, made at a depth of 30 feet on a bottom of rock and gravel
covered largely with Cladophora, aggregated about 50 yards in length. A quantity
of material was brought up and all critically examined. It was composed almost
wholly of red Chironomus larvae and their tubes, together with a few specimens of
Pisidium. A single ephemerid larva ( Ccenis) was the only other animal found. On the
various alongshore hauls, at or near the bottom and through the weeds, the following
forms were obtained: Larvae of Chironomus, ephemerid larvae, caseworms (a single
specimen seen), small hydrachnids, Physa and Pisidium (only a few of each), Cyclops
gyrinus and other species, Diaptomus sicilis var., Daphnella brachyura, Daphnia
arcuata, Eurycercus lamellatus, Acroperus leucocephalus, Leptodora hyalina, Bosmina
longirostris and some undetermined cyprids, Stylaria lacmtris and other annelids, and
an abundance of Hydra fusca of the brick-red variety. The most abundant thing was
Daphnella brachyura, and the next Diaptomus. The Cyclops was common, as were like-
wise ephemerid larv;e and the small bivalve entomostracans Eurycercus and Acroperus.

The towing-net collections, made in the open water some little distance from the
shore, contained a much smaller variety, all entomostraca and hydrachnids. The latter
were few in number, noticed only in a single haul. By far the most abundant species
were Diaptomus sicilis ( ?) and Daphnella brachyura, sometimes one predominating
and sometimes the other. With these were Leptodora and an occasional Daphnia.

The abundance of fishes in the lake was shown by the fact that our trammel net,
simply stretched in the open water in the evening and lifted at noon the following day,
contained 87 fishes, 12 of them trout, the remainder suckers and chubs, the latter most
numerous. At another setting of this net, near the mouth of Witch Creek, in 8 to 10
feet of water, 10 trout, 2 chubs, and 65 suckers were taken in six hours.

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 223

YELLOWSTONE RIVER SYSTEM.

Yellowstone River drains all the eastern and northern side of the National Park,
more than half its area, and from these waters much the larger part of our collections
was taken. Yellowstone Lake was visited both years ;, Pelican Creek and smaller
tributaries at the northern end were searched; and Yellowstone River was examined
at intervals from the lake to the mouth of the Lamar or " east fork." The smaller
tributaries of this system examined were Alum Creek, Tower Creek, Slough Creek,
Lamar Kiver, Amethyst Creek, Soda Butte Creek, Blacktail Deer Creek, Lava Creek,
Glen Creek, and Gardiner River. Collections were also made from numerous lakes
and ponds connected with this drainage system : Duck Lake near the west bay of Yel-
lowstone Lake; some alkaline ponds near Baronette's Bridge across the Yellowstone;
Fish Lake, near the Soda Butte; Twin Lakes, on the flat dividing the head waters
of the Gibbon from those of the Gardiner and drained by Obsidian Creek; Lake of the
Woods; Swan Lake, draining into Glen Creek; a small lakelet near Mammoth Hot
Springs, connected with the Gardiner; and Boteler Springs, outside the Park.

Yellowstone Lake. — With Yellowstone Lake we reach the aquatic headquarters of
this region, the real center of interest and importance for the study of the inverte-
brate life of Yellowstone Park. It is the largest lake so near the summit of the Rocky -
Mountains, and, excepting its high altitude, presents every feature suitable to the
maintenance of an abundance of animal life. Its zoological interest is fittingly sup-
ported by its geographic and scenic features, which supply an admirable setting to
the picture of life that slowly shapes itself in the mind of the zoologist as he studies
its waters and their contents and the inhabitants of its bottom and shores in their
relation to each other and to surrounding nature.

The geology of the region shows that the present lake is only the relatively small
remnant of a much larg er body of water which formerly filled Hayden Valley and
extended down the Yellowstone nearly to the present falls; but there is, I think, no
reason to believe that it has dwindled in zoological importance as it has in size.
Except for changes of climate, the variety of animal forms a lake of this size may
maintain ne,ed not be surpassed (and commonly is not) by that to be found in one
many times its size. It is not likely that there was ever here, when this lake was
largest and deepest, a special interior and deep-water fauna, such as occurs, for
example, in the Great Lakes of North America; for, if there were, remnants of it
would certainly continue and would have appeared in our deep-water dredgings.
As a home of animal life it has probably been for ages similar to what it is now,
except that we must suppose that the single species of fish which now inhabits it —
evidently an immigrant across the continental divide — has produced certain changes
in the balance of life, some of which will doubtless become more apparent as our
collections from this lake are thoroughly studied.

The most striking feature of Yellowstone Lake is the irregularity of its form and
the consequent length of its shore line, such that with an area of about 140 square

224 BULLETIN OF THE UNITED STATES FISH COMMISSION.

miles, its shore line approaches 100 miles in length ; a fact whose biological signifi-
cance will be understood if we call to mind that the greater part of the food of
fishes in such a lake is to be found among the weeds of its shoal waters alongshore.
From a main body (a fairly regular parallelogram, 7 by 12 miles) several large arms or
bays project to the west and south, giving the whole lake an extreme length of 20
miles and a width of 14. The immediate banks are generally abrupt, although not
high, with a narrow gravel beach, and except along the tributary streams, there is
little or no swampy ground tibout the shore. The Yellowstone above the lake, how-
ever, runs, as do most streams in this region, through a wide, swampy bottom, which
must afford an immense field for the breeding of fishes and for the growth of their food ;
and Pelican Creek, the next largest tributary, is similarly situated, while several of the
smaller creeks have each at their mouth a little bar, piled up by waves and ice, which
has partly stopped the outlet and so formed above it a weedy lagoon or marshy bay.

The bottom commonly slopes gradually downward, making an abrupt descent, so
far as known, only from the shores of Frank Island and from the lofty summits about
Southeast Bay. Beyond the beach of gravel and boulders which commonly borders
the lake, comes usually a belt of sand or sandy gravel, and beyond this a sandy mud,
becoming finer and darker inwards, until in the deepest water reached it was a very fine
black ooze. The greatest depth reported by the Hayden survey was 300 feet, in the
center of West Bay. My owu deepest soundings were incidental to our dredging oper-
ations, and were limited to a distance of 2,000 feet from shore off our camp at Hot
Spring Basin on West Bay, and to an area north and east of Stevenson Island, where
the lead was dropped at distances varying from half a mile to 2 miles from the island.
The greatest depth reached in this area was 231 feet, at a point nearly equidistant
from Stevenson Island and Steamboat Point.

My only temperature observations were made August 4, 1890, at which time the
suiface temperature of the water was 62° F., the bottom at 100 feet was 46°, and at
184 feet, 42£°.

This lake lies almost at the summit of the Bocky Mountain watershed, the con-
tinental divide following approximately the outline of its western and southern borders
for about 40 miles, at distances varying from a mile to 5 miles from the shore. To the
westward of the lake lie broken pine-covered hills, which rise from 250 to 800 feet
above its level. To the north are the dark ridge of the Elephant's Back, about 850 feet
above the lake, and the Sulphur Hills, which finally rise to a height of 9,000 feet above
the sea. On the east lies a mass of rugged volcanic mountains, a part of the Absaroka
Range, patched with snow all summer. They approach the shore most closely along
the southeast arm at the upper end, where the scenery is very bold and fine. A few
peaks rise to a height of 11,000 feet above the sea at a distance of 5 or 6 miles from
the shore. The boldest elevations are those just below the mouth of the inlet, where
mountains less than 2 miles away reach a height of 9,600 feet.

Besides the Upper Yellowstone and Pelican Creek, already mentioned, the principal
tributaries to the lake are a number of small streams which drain these eastern moun-
tains, and, taking their waters from the melting snows and flowing most of their way
through overhanging forests, bring to the lake a considerable amount of very cold
water. The hot springs and geysers are found mostly on the western arm and at
Steamboat Point in the northeastern part of the lake, but the amount of warm water
contributed by them is quite insignificant for so large a lake.

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 225

In the open water there was always a very fair supply of entomostraca, both .
Cladoccra and Diaptomi, .but at the time of our arrival on the West Bay the phenom-
enal fact was the vast abundance, both in deep and shallow water, of & rotifer or
wheel animalcule which forms rolling spherical colonies imbedded in a gelatinous
medium, each colony consisting of a little cluster of these animalcules arranged in
such a man uer that their inner ends approach each other in the middle of the mass,
while their outer ends, with mouths, cilia, etc., are exposed on the surface. To the
naked eye these colonies of rotifers appear like minute grayish specks of floating
matter. This species belongs to the genus Conochilus, but differs noticeably from the
common C. volvox. I have thought best, consequently, to describe it as G. leptopus
(page 256). It was so abundant in the water that a haul of a ring net, a foot across,
for fifty strokes of a single pair of oars gave a measured half pint of this form alone.

This colonial rotifer is not to be confounded with the "water bloom," which devel-
oped in Yellowstone Lake a little later to an extent very embarrassing to our suri'ace
net work. This so-called " bloom " consisted of specks of various alg;e growin g so freely
in the water as to give it a faint tint of dirty green, and washing ashore in quantity
along the leeward side of the lake, usually at this season the northern and eastern.

Away from the shore, by far the most common crustacean was Daphnia pulex*
Although in ordinary situations the males of Daphnia are by no means common,
in our Yellowstone Lake collections, made in August and September, the males of this
variety were many times commoner than the females, making sometimes nearly the
whole of a large catch. The few examples of the other sex seen were mostly young,
although a female bearing the ephippium occurred occasionally. Next in abundance
was the smaller of the Diaptomi found in Shoshone Lake, the so-called variety of D.
sicilis, and with this came somewhat rarely, but still fairly abundant, D. shoshone and
D. lintoni. Several species of Cyclops occurred here, only a new one ( C. minnilus)
very frequently, however, and this in small proportion.

Most of these crustaceans ranged in shore as well as in the deeper water of the
interior parts of the lake, Daphnia pulex falling away in numbers more rapidly in
shallow-water collections than Diaptomus sicilis. To these inshore species we may
add, from our surface-net collections, Polyphemus pediculus (sometimes very abundant
among the weeds), Cyclops gyrinus, and C. serrulatus (both rare), Chydorus sphcerivus
(few), Scapholeberis mucronatus (few), Cypris sp. (only occasional), Alona, and the usual
miscellaneous drift of shore forms, Chironomus, Allorchestes (both dentatus and inermis),
Oammarus, caseworms, hydrachnids, planariaus, Clepsine, etc.

Collections made in the lake near enough to the outflow of hot springs to exhibit
their influence differed from those made in cold water only in their more scanty char-
acter; and where the water was actually warm it commonly contained nothing but the

"The common anil oven abundant occurrence of this species in Yellowstone Lake as a form
apparently pelagic in its habits (widely contrasted, consequently, with its usual character) was so
unexpected and unusual that I hesitateil long before assigning this Daphnia to the species most abun-
dant in our stagnant pools. Prolonged study of it from various collections in the Park in comparison
with those from the waters of Illinois, has finally led me to conclude, however, that this Yellowstone
Lake form is not to be specifically distinguished from American examples of pulex. In order to
furnish material for a more critical comparison than has hitherto been made of the American and
European representatives of this species, I append a description, under the varietal name of pulicaria,
based upon Yellowstone specimens, with figures of both sexes (page 242 and plate xxxvn, fig. 1).

F. C. B. 1891 15

22() BULLETIN OF THE UNITED STATES FISH COMMISSION.

larger crustaceans, insect larvse (especially caseworms), and other alongshore material,
together with the dead and empty debris of insect transformation. It was in fact
very clear that the frequently observed basking of small fishes in these warm waters
was not caused by a greater abundance of their food.

A comparison of the collections made at and beneath the surface, by day and by
night, iii sunshiny and in cloudy weather, would seem to indicate that the lake variety
of Daphnia pulex is much more sensitive to sunlight than any other associated form.
In collections made at the surface after dark, and in those made in sunny weather
below the surface, this was many times the most abundant crustacean ; but in similar
collections made at the surface in sunshine it was relatively rare, Diaptomus sicilis
then taking the lead.

Only brief time and scanty opportunity could be had for miscellaneous alongshore
work during our visits to Yellowstone Lake. The greater importance, from our point
of view, of deep-water work, the stormy weather, and the unfavorable character of the
beaches, commonly either covered with bare gravel or packed with large boulders and
beaten by surf, made such work unprofitable and difficult. Nearly all our knowledge
of this alongshore fauna we owe, in fact, to a day spent in 1891 on Stevenson Island
(1J miles long by £ mile wide), lying 2 miles from the hotel landing, in the north
end of the lake. The shores of this island vary from precipitous bluffs on the west to
a weedy shallow on the south. The beach is gravelly everywhere, except as boulders
thickly cover the bottom and banks, and outside this is a belt of gravelly sand, fol-
lowed by sandy mud still farther out. Water weeds occur in scanty patches, chiefly
Chara and algae ; a coating of minute, dirty alga>, commonly covers the stones. On the
stones are also countless tubes of small Ghironomus larvae, almost covering the sur-
face, mostly emptied by the maturing of these insects at the time of our visit. Under
the stones are a considerable number of leeches of various sorts (mostly Nepheiis and
Clepsine), and an occasional small annelid worm. On them, among the weeds, a small,
black, spiral mollusk, Physa, may occur by hundreds, and creeping under or over them,
or swimming through the water just above the bottom, we found an occasional Gam
marus. A large species of Gorisa, several water-beetles (Deronectes), and a perlid
larva were all the other insect forms taken here. A small collection of entomostraca
from the weeds has not been determined.

Our knowledge of the bottom fauna of Yellowstone Lake is based on the product
of eleven dredgings, at depths varying from 15 to 198 feet. Four of these dredgings
were in shallow water, 20 feet or less ; two were from a medium depth (25 to 50 feet) ; one
ranged from shallow water to deep (20 to 120 feet); one from a medium to a consider-
able depth (40 to 100 feet) ; and three may be classed as deep throughout, ranging
from 186 to 198 feet. In this deepest water the most abundant inhabitants of the
bottom were long and slender annelid worms ( Oligochceta) not yet studied, blood-red
larvae of Ghironomus of considerable size, and a small bivalve mollusk (Pisidium).
Several leeches also occurred in our deepest hauls (Nephelis maculata and species of
Clepsine), a few specimens of Gammarus, probably taken after the dredge left the
bottom, several small planarians, a lot of living Spongilla, and in one instance a Phyaa on
living vegetation — in all probability a floating fragment, as the haul on the bottom
here was made at not less than 186 feet. In other dredgings, at 100 feet or more, many
specimens of Cypris, a very few hydrachnids (probably swimming above the bottom),
and several worms (uematoids) were added to the above list.

. AQUATIC INVERTEBRATE FAUNA OP WYOMING AND MONTANA. 227

As we worked into shallower water the Gammari and the leeches became more
abundant, especially Glepsinc, and at '25 to 50 feet univalve in.ollu.sks (Planorbis exacutus
and Physa), Allorchestea inermis, small white Chironomi, larvae and pupae, and case-
worms in sand tubes appeared. At 15 and 20 feet, among the weeds, the assemblage
of associated animals was Daphnia, Diaptomus, Conoohilus, all abundant, many oligo-
chajte worms, leeches and leech capsules, Physa;, Limnaxe, Allorchextes, Gammari,
and cyprids (Cypris barbatus, n. sp.), various caseworms, nymphs of Ephcmeridce,
Chironomus larvae, and larvae of Tabanidte and Culicidce.

No discussion, of the zoological resources and relations of Yellowstone Lake with
reference to fish-culture would be even approximately complete which ^did not take
account of the animal contents of the streams and other waters connected with it,
since these are the principal resorts of the young of the only species of fish the lake
now contains and must always be the chief breeding-places of fishes generally.
Apart from the river above and below the lake, the most important tributary is Peli-
can Greek, a peculiar stream for a mountain region in the fact that for 2 or 3
miles of its lower course it is broad, muddy, and comparatively sluggish, more like a
bayou than a creek, thick with vegetation, and much frequented by water birds,
whose feathers floating on its semi-stagnant surface gave it the appearance of a barn-
yard pond. Above this stretch, although still bordered by willow-covered and more
or less marshy bottoms, it becomes swift and rocky, except where cut across by
numerous beaver dams.

The sluggish waters just above its mouth are, as might be expected, rich with
small crustaceans aud insect larvae. Amphipod crustaceans were very scarce in
this creek, Gammarus not occurring in our collections, and Allorchestes but once; and
among eutomostraca, gigantic specimens of Eurycercus lamellatus were far the most
numerous, making three-fourths of the entire bulk of the product of hauls made in
open water and among alg;e and other water weeds. With these, in open water, were
a new species of Macrothrix, a Diaptomus, an occasional Daphnia pulex, several speci-
mens of Cyclops, a few of Daphnella aud of Cypriit, and a single Allorchestes. Among
the algie, besides the foregoing, several examples of Simocephalus vetulus were taken,
together with Bosmina, Ceriodaphnia, Polyphemus, and Chydorus sphwricus. The most
abundant insects were, of course, Ghironomi — larva}, pupae, and adults just emerging —
and ephemerid larva}. A few Goriscc aud caseworms, some small aquatic Coleoptera,
and a single living Limncea were also noticed.

In the shallower and swifter parts of the stream insect larva? take the lead, the
bulk of the collections consisting of large and small caseworms of various species,
most of them attached to stones, larvae of Ghironomux, ephemerid and perlid larvae,
large Gorisce, and several beetles of small size. The caseworms and ephemerid larvae
were exceptionally common.

On the whole, this stream — which must stand for the present as an example of
many others — contained invertebrate forms of animal life in very fair abundance; in
the swifter waters the insect larvae (neuropterous and dipterous) which lurk under
stones, and in the more quiet parts entomostraca and insect larva- of different habit.

A much smaller creek, known on the map as Bridge Creek, and noted among
tourists because crossed by a perfect and highly picturesque " natural bridge," has at its

228 BULLETIN OF THE UNITED STATES FISH COMMISSION.

mouth ;i small pond or lagoon which was searched with some care during half a day.
This pond is shallow and muddy, but mostly clear in the middle, with a fringe of
aquatic vegetation and stout marsh grass growing in the water. The clear water was
full of minute spherical masses of alg;e, among which were a Diaptomus and Poly-
phemus— apparently the usual lake species, but not preserved. In the grass were great
numbers of Oammnrus robustus and a few Allorchestes iiiermis, and ephemerid larvae
were abundant everywhere. Chironomus larvae were common in the collection and
doubtless very abundant in the mud, and a robust Gorethra larva was occasionally
taken. Small water-beetles (Deroneutes and Hciliplun), caseworms with cases made of
fragments of vegetation and others of line gravel, Corisa, a water-skater (Hyyrotrechus),
black Podutidcv, and occasional terrestrial insects were among the other insect elements
available for the food of fishes. The mollusks of this little collection were small
LlmncBce, Physfc (large and small), Pisidium, and Valvata. Various leeches (the most
abundant the common Nephelis maculata), Hydrachnidce, and plauarians complete the
list thus far made up.

The pond was swarming with young mountain trout (Salmo mykiss), a few of
which I dissected for a determination of their food. One of these an inch and a half
in length had eaten Ghironomun larvse and imagos chicHy, the remainder of its latest
meal consisting of other insect larvie not. in condition to identify, and the entomostra-
can Polyphemus pcdiculus. A second, an inch and a quarter long, had also fed mostly
on Ghironomux in its various stages of larva, pupa, and imago, but had made about
a third of its meal from entomostraca (Daphnia pulex and Polyphemus ped,iculus).
Another, still smaller (0.92 of an inch long), taken from the open lake, among the small
weeds growing on a flat, muddy rock, had filled itself with Chironomus pup;o only, as
had still another of the same size. A third specimen from this situation had eaten
more larva* of Simulium than of Chironomus, and a fourth had also eaten Simulium
Iarva3 and another dipterous larva unknown to me.

I may add here that other young trout, in a small, swift rivulet near the Lake
Hotel, were feeding continuously, August !>, on floating winged insects, mostly, if not
all, Chironomus and smaller gnat-like forms.

The large leeches at Bridge Bay (Nephelis maculuta) betrayed their scavenger
habits by collecting in numbers upon a dead fish, which they were evidently feeding
from. Two specimens taken elsewhere in this pond proved on dissection to have the
alimentary canal nearly empty, one containing only a few fragments of Gammarus in
the rectum, and the other a single leg of a Gamm/irus in the oesophagus.

As illustrations of the smaller animal life of the river below the lake and in its
vicinity, I may report the product of two trips made August 11 and 23, one from a
quarter to half a mile below the outlet, and the other to a point about a mile below.

The most fruitful ground at the first locality was a sedgy flat on the left bank
and a bed of flat rock covered with algaj and other fine vegetation, with about 6 to 8
inches of water. Other collections were made from the bare sandy bottom, in water 6
inches deep, with moderate current.

On the weedy rocks occurred the large hairy Gi/pris described herewith (p. 244) as
G. barbatus, and other small blue cyprids not yet studied. The presence of Ghirono-
mus larvae, several sorts of caseworms, larva) of ephemerids, Hygrotrechus, Gorisa (larva
and adultl, various water-beetles, specimens of Gammarus robustus and Daphnia pulcx,

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 229

Physa (small ami large, ia quantity), Pisidium, many Limncea;, of various sizes, Planor-
bis, and small Amniuolce, accounted sufficiently for the fry of the mountain trout
abundant among the woods. The mast interesting object here, however, was a small
cylindrical brown turbellarian, which lias the habit of swimming freely through the
water, rolling over steadily from side to side as it swims, Although common here and
easily taken, every effort at its preservation failed completely, the specimens going
to pieces in spite of the varied use of hot water, corrosive sublimate, cold and hot
osmic acid, Pereuyi's fluid, etc. This interesting form seems, notwithstanding, worthy
of special mention, and I have drawn up the following brief description, made from
field notes, which may serve to identify it to some collector more fortunate than I in
his opportunity to study it closely.

Form cylindrical, tapering a little toward both ends, the posterior end blunt-
pointed, the anterior flattened in creeping, and broadly rounded. Locomotor surface
not specially flattened. When swimming the two ends are similar. Length, when
extended, 5 to 9 millimeters, width 1 millimeter. To the naked eye dark reddish or
orange, slightly paler before and beneath. Closely examined, the color is in minute,
irregular flecks on a yellowish ground, and varies in intensity, of course, according
to the extension of the worm. Sometimes the intestine shows through as a darker
median shade, and the orange-brown eggs, 0.25 millimeter in diameter, also deepen
the color locally. When emptied of these, its color is nearly uniform reddish-brown.
The eggs are spherical, conspicuous, in two ovaries, one each side of the abdomen,
and, to the number of twenty, may nearly fill the body. A pair of eye-spots placed at
a distance from the front of the head about equal to the diameter of the body.

These worms were found in 1891 (September 1), much more abundant than at
the above locality, in some clear, gravelly pools filled with filamentous alga; along
Soda Butte Creek. They were everywhere thick among the algae, and could be col-
lected by scores in an hour.

At the lower locality mentioned above, several ephemerid larvae, specimens of
(jammarus robwtita and of Allorchentes inermis, caseworms with cases made of fragments
of bark, larv;e of Simulium, and some small planarians were found.

Finally I close this preliminary account of our Yellowstone Lake collections by
noting the results of a brief examination of the contents of the warm waters along
the shores of the hot-spring basin of West Bay.

Hauling August 3, 1890, in shallow water only a few feet from shore, at tempera-
tures varying from 70° to 101° F., where the ordinary surface temperature was 62°,
we took a great quantity of the rotifer GonoeMIus leptopus, very many examples of
Polyphemus pediculus, a few Diaptomi, and a very few specimens of Daphnia pulex.
There were probably five times as many examples of Polyphemux as of all other ento-
mostraca. The Diaptomi were D. shoshone (a few) and several I), nicilis, and all the
other cntomostraca were a few each of Cyclops xerrulatus, Saapholcberis mucronntus,
and Chydorun xphtcricm. There were no insects in these collections, living or dead,
and the total amount of animal life was much below that of the cold water adjacent.
In a spring near shore, witli a temperature of 103° F., containing much Ottdllaria and
full of a dark-red alga, there were many holotrichous infusoria and other smaller ciliata,
minute flagellata, a fine anguillnlid, a small, active planarian, and many examples of a
rotifer (Monostyla) allied to M. cormtta, but apparently new. (See page 250.)

230 BULLETIN OF THE UNITED STATES PISH COMMISSION.

Duck Lake. — As additional material for comparison, the results of a visit to Duck
Lake, close beside the West Bay of Yellowstone Lake, may be worthy of present men-
tion. This clear, cold lakelet, about half a mile long and three-fourths as broad, lies
in a steep, oval hollow of the woods, its shore without beach, too deep for vegeta.tiou,
and surrounded by a tangle of fallen trees — a secluded woodland pool. The special
peculiarity of the little collection brought in from here August 4 consists in the pre-
dominance of Diaptomus lintoni (the only Diaptomus found) over the other entomos-
traca, and the vast abundance of a shelled infusorian (Difflugia glolnilosa) brought up
by the dredge at 65 feet and by the towing net sunk to a depth of about 30 feet.

Alongshore a small but miscellaneous collection was made of dragon-fly larvae
(Libellulinw and Agrionina;), larval May flies and Chironomus, of several AmpMpoda
(mostly Allorchestcs dentata), of various entomostraca, among which were Simoceph-
alus vetulus, Cyclops gyrinus, and other species of Cyclops, Alona, etc., and of the
common large leech, Nephclis maculata. In the dredge, besides the Difflugia already
mentioned, were several specimens of cyprids (Candona), very many Chirononnis
larvae in their tubes, Cyclops minnilm (a few), a few Corethra larvae, Diaptomus -lin-
toni, and a small auguillulid worm. The water at some depth was loaded with small
pellets of uniform size and similar shape, made up of diatoms, fragments of filamentous
and other algaj (mostly emptied of chlorophyl), and of other vegetable debris together
with grains of sand, all of which had the appearance of being the excrement of the
common Ghironomm larva. So thick was this material that it soon lined the surface net
when hauled some 30 feet below the surface. With it came, besides Difflugia, a few each
of Corethra larvae, Daphnia pulex, Diaptomus lintoni (all females or young), and a single
Gammarns. A surface haul gave a substantially similar product, with the addition
of the entomostracan Sida crystallina, not recognized in the adjoining lake.

Lake of the Woods. — Here, as. well as anyhere, may be reported the product of a
very little work done with the dip net along the margins of a little oval pond a quarter
of a mile in length, lying among the hills above Obsidian Cliff, at a height about the
same as that of Yellowstone Lake. It has neither outlet nor inlet, and is doubtless
fed -by springs. It was evidently shallow, although it was not sounded by ns, its
bottom apparently fathomless mud, and the open water of its center bordered all
round by a belt 100 feet wide of pond lilies and the usual accompanying vegetation.

Collections could only be made among the lily pads with a hand net from a log
near shore. They were remarkable only for the variety of entomostracan and insect
forms and the vast abundance of a Stentor which blackened the surface in patches
some inches across and covered the lower surfaces of the lily pads as if with a layer
of soot. This is allied to Stentor igneus, from which it differs, however, by characters
to be derived from the description published on page 256. The principal insects taken
were ephemerid and Chironomus larvae, a few caseworms, a specimen of the water-
beetle Oraphoderes fasciaticollis, many black spring-tails (Poduridw), and several
water-spiders. Sida crystaUina was the most abundant crustacean, but specimens
were also taken of Scapholeberis mucronatus, Cyclops, Diaptomus, Simocephalus vetulus,
and Acroperus leucocephalus. A few examples of Allorchestes dentatus were also seen,
aud a fragment of a hairy, bristled worm (Naidomorpha).

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 231

GARDINER RIVER SYSTEM.

Turin Lakes. — Collections were made August 20 from the upper of two small,
closely connected lakes called the Twin Lakes, lying in the boggy trough between the
hills beside the main road from Mammoth Hot Springs to the Norris Geyser Basin.
This and the two following lakes belong to the Gardiner Eiver system. The Twin
Lakes give origin to a small stream known as Obsidian Creek, through which they
are connected with a great expanse of swamp and shallow weedy water, known as
Beaver Lake. The upper lake is a clear, clean-looking pool, with much marginal
vegetation (lily pads and the like), and with boggy banks which drop off suddenly,
forming an overhanging grassy margin. Several discolored springs open into the
lake, discharging into it water which is said to contain alum in solution.

The dredge brought up from the deepest water found (beginning at 30 feet and
ending at 39), a quantity of very soft, streaked, ill-smelling mud, with a little dead
vegetation and a very small proportion of animal life. This consisted mostly of Chiron-
omus larvae, partly red, but most of them faded brown, as if discolored by their
surroundings. The only other product of the dredge was two specimens of Gamma-
rus, a single leech (Clepsine), and one Pisidium. The water itself, however, was well
stocked with animal life, and a haul of a towing net above the bottom, at a depth of
30 feet, at 11 a. m., in bright sunshine, with a stiff breeze blowing, gave a consider-
able number of Gammarus, a very good collection of the characteristic entomostracan
of this lake (Diaptomus lintoni), and several specimens of Daphnia and Corethra larvae.
A surface haul under the same conditions gave a few examples of Daphnia schcedleri,
an occasional Cyclops, a single ephemerid larva, and a large quantity of Diaptomus
lintoni. Alongshore, upon the weedy bottom — an admirable lurking and feeding
ground for lish — were the commoner insects (Notonecta, Hyyrotrechus, ephemerid and
agrionine larvae), several specimens of Gammarus, a great quantity of the eutomos-
tracan Sida crystallina, and a few Simocephalus vctulus and Chydorus. Curiously, not
a caseworm was taken from this lake — a fact possibly to be explained by the peculiar
character of its bottom. A careful search was made from the boat and along the
bank for signs of a plant of monntaiji whiteflsh made here the preceding year by Mr.
Lucas of the U. S. Fish Commission, but no trace of them was found.

Swan Lake. — This lake, a quarter of a mile long by two thirds as wide, is of nearly
the same size as the two preceding, but is, perhaps, the shallowest of all (not over
3 feet in depth). It lies on a plateau of the same name, not far beyond Terrace
Mountain and beside the main Hot Springs and Geyser Basin road. Its waters are
derived from the adjacent mountains to the west, and pass out through Glen Creek into
the Gardiner. As it lies in a plain, its immediate surroundings are level. Its bottom
is of rock and sandy mud, with Chara and other weeds, and a strong growth of rushes
inshore.

The collection lists from this little lakelet are unusually full, a fact apparently
due chiefly to its geological surroundings. All the waters previously discussed are
situated in the Park plateau, and the rocks of their drainage basins are all lava in
some form, usually that modification of it known as rhyolite. Swan Lake, on the
other hand, is in a cretaceous region, where the geological deposits are largely lime

232 BULLETIN OF THE UNITED STATES FISH COMMISSION.

stone. We found here, August 29, seemingly in consequence of this fact, an extra
ordinary abundance of mollusks; large Limnwa; in the grass at the margin, a small
Planorbis, Pisidium, Sphwrium, etc. Gammarus occurred here under stones, and
numbers of Allorchestes dentatus were taken in the weeds.

The insects collected were species of Hydaticus and Hydroporus, Notonecta and
Corisa, caseworms, and Chironomus larva?. A few water-spiders (hydrachnids) were
taken among the weeds, and on the stones were great numbers of the cocoons of the
large leech, Nephelis maculata. Young of this leech were numerous, and an occasional
adult was seen, with specimens of Clepsine ornata and other species of the genus.
The most abundant entomostraca were Daphnia schoedleri and Diaptomus siciiis, var.,
frequent specimens of D. shoshone occurring with the latter. That this large species,
previously found only in the larger lakes, should replace here in this shallow pond the
Diaptomus lintoni elsewhere characteristic of shallow water, was another of the sur-
prises of distribution and association of which these mountain lakes are fruitful. The
other small Crustacea noted were Cyclops, Eurycercus, Alona, and several species of
Cypris. The occurrence of tipongilla here is also worthy of mention.

Oardiner Lakelet. — Similar to the foregoing in -geological situation, but smaller,
deeper, and surrounded by deep and broken hills, is. a little pond formed in the course
of a swift and rocky stream to the west of Gardiner River, between Mammoth Hot
Springs and the town of Gardiner. It is about 250 yards long by three-fourths as
wide, and 20 feet deep in the interior, with its marginal waters filled with a strong
growth of rushes and other vegetation. It was full of eutomostraca, gammarids, and
insects. On the stones were great numbers of the egg capsules of the common large
leech, Nephelis maculata, and many young of this species were crawling about, but
the time for the adults had apparently gone by. The assemblage of entomostraca
was again peculiar, the most abundant form being a Ceriodaphnia, and the next
commonest a medium-sized Diaptomus (mostly immature), described on page 253 as
a new species under the name of />. piscince. There were also many specimens of
Cyclops, an occasional Daphnia pule.v, and several of Ghydorus, in a collection obtained
by drawing the surface net through the open water. Chironomus lar vre were, of course,
abundant; and worms of various kinds, small flat planarians, larvae of Agrion and
Dytiseidm, species of Haliplus, Corisa, and Notonecta, larvse ofLibellulw and ephemerids,
species of Clepsine and of Physa, were represented in our collections.

Small ponds. — Occasional roadside collections made from small standing pools
will be of interest chiefly because of the locality and altitude. The ordinary contents
of such waters at the time of our visit may be learned from the following lists:
Weedy pool between Norris Basin and Grand Canon, August 21, 1890:

Chirouomus larvae (a few).

Corethra larva (one).

Corisa larvae (three).

Diaptomus lintoni (many).

Cyclops (a few).

Polyphemus pediculus (a few).

Daphnia pulex (many, with ephippia).

Daphuia schoedleri (a few).
Scapholeberis mucronatus (very many).
Ceriodaphnia reticulata (many).
Chydorus spha'ricus (very many).
Macrothrix sp. (one).
Acroperus leucocephalus (one).

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 233

i
Woody beaver pool near Soda Butte Creek, September 1, 1891 :

Agabus sp. (two).
Hydroporus sp. (several).
Colymbetes sp. (one).
Dytiscidne (larvae).
. Corisa sp. (several).
Ephemerid larvae (a few).

Chironoraus larvne (many).

Phryganeida- (many, with cylindrical tubes made

of cylindrical pieces of vegetation).
Turbellaria (brown cylindrical species. See page

229).

Weedy poud near Soda Butte Creek, September 1, 1891:

Chironormis larva; (a few).
Corethra larva (one).
Ephemerid larvso (several).
Cyclops sp. (one).

Simocephalus vetulus (very many).
Ceriodaphnia sp. (very many).
Daplmia pulex (many).
Annelida (one, fragment).

Standing pools left by Soda Butte Creek ; covered with algae ; muddy bottom ; Sept. 1, 1891 :
Hydroporus sp. (one). , Cyclops sp. (one).

Chironomus larvae (a few).
Ephemerid larva; (many).
PhryganeidiB (small, one).
Podurid (great quantity on algie).
Gammarus (one).

Daphnia schncdleri (many).
Simocephalus vetulus (very abundant).
Ceriodaphnia reticulata (many).
Physa sp. (large, two).
Chaatogaster «p. (one).

Two circular ponds, each ap])roximately 500 feet across, situated about a mile
from Barpnette's Bridge and beside Lainar River, were examined in passing, so far as
could be done by alongshore work and by wading out with a surface net. In one
there was an abundance of vegetation — rushes and a variety of other water weeds —
and no appearance of alkaline deposit, the bottom being a film of mud on gravel. In
the other there were no rushes, but the water was sufficiently alkaline to have a smooth
feeling, and the dead water weeds were whitened as they lay upon the bank.

In the flrst pond, there was a great quantity of dead shells of a large PlanorMts,
and fewer of a large Limncea around the margin. Vast numbers of Allorchestes dent at a
occurred on the vegetation, and especially in the soft mud of this pond. The ento
mostraca were nearly all Copepoda, of the genus IHnptomm, most of them T>. lintoni.
Not a single Cyclops was noticed, nor a single Daphnin. A few Ceriodaphnia; occurred,
several specimens of Simocephalus vetulus, a very few of Chyditnut spluerictut, and, for
the rest, a considerable number of hydrachnids, a few Chironomus larva;, and several
larvse of dragon-flies (Agrion). With the foregoing were the common large Corisa
of this region, Notonecta, Physa, Deronectes, and a small hydrophylid larva.

In the alkaline pond near by were a very few mollusks and a moderate number of
insects, the latter consisting chiefly of agrionine larvse and small larvie of Chironomns.
The entomostraca were much as before, except that Diaptiimus shoshone in small
numbers mingled with D. lintoni.

234 BULLETIN OF THE UNITED STATES FISH COMMISSION.

MADISON RIVER SYSTEM.

Although large collections were made from streams of various sizes whose waters
find their way into the Missouri by way of the Madison, the only lakes so connected
upon which we worked were Mary Lake and Grebe Lake, the former draining through
Nez Perce Creek into the Firehole, and the other giving origin to the main branch of
the Gibbon.

Mary Lake. — This little lakelet, situated at a level of 8,200 feet, on the divide
between the waters of the Firehole and those of the Yellowstone, is an oval body of
water some 5 or C acres in extent, clear but shallow, with a fine gravelly beach
and very little vegetation. The immediate banks are rather steep, and the country
around is densely wooded with pine. The greatest depth at our visit was only 20
feet, and a lagoon-like bay near the lower end was but 5 or 6 feet deep. Both inlet
and outlet — the former very small — were dry, but the lake overflows into Nez Perce
Creek when the snow goes off in spring.

Our collections from this lake were particularly remarkable for the great number
of one of the most beautiful and interesting of our fresh-water entomostraca, the
species known as Holopedium gibberum. In a haul made with the towing net at the
surface, in the shallow water near the outlet, a great quantity of this species was taken,
together with a still greater number of Diaptomm lintoni and an occasional Corethra
larva. The sun was shining at the time (11 a. m., August 14) and the water was
rippled by a light breeze. Near the bottom, at a depth of 16 feet, D. lintoni was the
prevailing form, mingled, however, with an almost equal quantity of Corethra larva;
and a considerable number of Daphnia schcedleri.

The deeper bottom was of sandy mud, which contained a large number of Chi-
ronomus larvre in tubes — most of them the common large red species — a great number
of the common form of Pisidiwm, and many caseworms with tubes composed of sand
grains, several of them bicarinate. The dredge brought up a little Spongilla, several
of the usual annelids, and Corethra larvje, but no living vegetation. No amphipods
were taken from the lake, and no univalve mollusks. An unusual number of aquatic
insects occurred alongshore, most of them specimens of Agabus (two species) and
Deronectes griseostriatus. The leeches were, as usual, several species of Clepsine and
Nephelis maculata.

Grebe Lake. — This shallow, muddy lake, connected with the head waters of Gibbon
River, is so far secluded within the forest that it has no current name, and is locally
almost unknown. We reached it August 27, with saddle and pack animals, from the
Grand Canon Hotel, carrying canvas boat, seines, and our smaller apparatus.

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 235

Our' collections were made with a surface net in deep and sballow water, with a
dredge at a depth of 36 feet (the greatest found), and with the hand net from grass
and lily pads near the margin, from the gravelly bottom in shallow water, and from
weedy mud — this list exhausting, in fact, all the varieties of situation offered.

The lake lies north and south in greatest length, the outlet leaving the south end
and flowing to the south at first. It is in some respects a duplicate of Mary Lake,
but is somewhat larger, being half to three-quarters of a mile long and about two-
thirds as wide. It is of oval form, with grassy margins, commonly sod to the water's
edge, rimmed round with lily pads and other water weeds, and with a bottom of soft,
black mud. The banks were somewhat swampy, but the ground was higher to the
north and west. Three small streams now into the lake, one from the northeast and
two from the west.

Although so unlike Shoshone Lake, its assemblage of animal forms bore a striking
resemblance to that of the larger, clearer lake. The absence of fish, the abundance
of Oammarus and D-iaptomi, and the scarcity of Daphniida; are examples. On the
other hand, the grassy borders and weedy shallows entertained a much greater abun-
dance and variety of insect forms than the hard and bare margins of either Shosho ie
or Mary Lake.

In the mud of the bottom were many large red Ghironomus larvae, a few speci-
mens of Gammarus, and the usual Pixidium. The entomostraca were mostly Dinpto-
mus lintoni, which replaced in this small lakelet the J). shoxhone of the other lake; here
also we found Daphnia clathrata, n. s., the only locality thus far discovered for it. It
was not abundant in Grebe Lake, and may have bred primarily in the swamps adjoin-
ing. A species of Cyclops also occurred here in small numbers, which is described on
page 248 as C. capllliferus.

Inshore collections were unusually fruitful. Q-ammarus and Allorclic'stc.i were very
abundant along the margin in the weeds and grass, and Pisidium especially was
extraordinarily common. Here also were agrionitie and ephemerid larva}, caseworms
with cases of fine sand, Corim, Piaidium, Phym, Chironomns, and Flpongilla, and on
the mud among grass and algra were dytiscid and small sialid larv;e, Phi/so, and Pisid-
ium, Nephelis and Glepxine, Allorchestes and Gammarus, Ghironomus and ephemerid
larvae, larvse of dragon-flies, and specimens of Haliplus.

This lake was an additional illustration of the fact that, in this high mountain
region, where aquatic life seems oppressed with unusual difficulties, change in cir-
cumstance takes extraordinary effect, so that each lake has its distinct and special
zoological character.

236 BULLETIN OF THE UNITED STATES FISH COMMISSION.

FLATHEAD RIVER SYSTEM.

The waters of the Fhithead region from which collections were made by us were
Swan and Plathead lakes and Flathead, Swan, and Coiur d'Alene rivers, and the
Jocko at Kavalli. Those from the lakes only can be here discussed.

Flathead Lake. — Although this lake stands in some respects in decided contrast to
Yellowstone Lake, these differences tend largely to neutralize each other. Flathead
Lake is over 200 miles farther northward than Yellowstone, but the latter is 4,775 feet
the higher above the level of the sea. These lakes lie on opposite continental slopes,
their waters passing respectively into the G-ulf of Mexico and the Pacific Oceau, but
neither is more than a few miles from the relatively low continental divide, easily
passable by most of the plant and animal forms likely to occur in such waters. Both
lakes lie in the course of streams of considerable size, but these streams flow iu
opposite directions, the inlet of Flathead Lake coming southward from the British
Possessions, and its outlet running first to the south and then to the west as Flathead
Kiver, a branch of the Columbia, while Yellowstone River, rising about 50 miles
above the lake, runs northward more than a degree below it before swinging to the
east to join the Missouri. Nevertheless, the headwaters of the two river systems inter-
lace almost inextricably through interlocking mountain valleys along several hundred
miles of the main Kocky Mountain range. Both lakes lie among mountains from
whose rugged gulches the snow never wholly disappears, and both are bordered by
forests broken by park-like openings on the lower slopes; but the geological structure
of the surrounding country and the chemical composition of the rocks which form
their shores and beds differ widely for the two, and the forests, all pine and fir and
other conifers around Yellowstone Lake, are largely deciduous trees about Flathead.

The lakes are similar in size and are both deep enough to give a deep-water
character to their interior fauna, but Flathead has much the more uniform shore-line
and contains — if I may judge from the parts of it which we examined — a larger
extent of shallow and weedy water. It is divided, iu fact, by a chain of islands
stretching across its lower third, into unlike parts ; the northern deep and clear, and the
southern shallow, and easily stirred up to its clayey bottom by the winds.

Finally, both lakes, like most of this region, are evidently far smaller now than
they were in an earlier geological period. The extension of the old Flathead above
the present lake is shown by the terraces marking its former shores, which may be
traced, one above the other, for a considerable distance above the inlet; while Hayden
Valley, the deserted part of the Yellowstone Lake, lies below the lake along its pres-
ent outlet.

The Flathead is reported by steamboat men and residents to be about 25 miles
long by 10 or 12 wide, although the best published map of the region makes it 24
miles long by 17 wide; but as the country about has not yet been surveyed, neither
distances nor proportions are precisely known.

The immediate surroundings of this lake are attractive in the extreme. Beside it
on the east lies the Mission Kange of mountains, beginning to rise almost from the water's
edge, and presenting to a near viesv, along the lower half of the shore, a curiously reg-

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 237

ular series of high, scantily wooded ridges and rocky gulches transverse to the length
of the lake. Further back the peaks of the higher mountains rise bare and steep.
This Mission Range diminishes in height northward, and falls away to Swan River,
near the northeast part of the lake, but across the river to the east and north the
Kootenai Range continues far up along the Plathead. Opposite Mission Range, on
the western side of the lake, lies a mass of heights between mountain and hill, rising
one above another, mostly wooded, but with occasional park-like openings. Above the
lake a level valley several miles wide, partly densely wooded and partly prairie, extends
above Kalispell, and to the south lies the naked plateau of the Flathead Reservation.

The principal tributaries are the Flathead, a still, broad river, larger than the
Yellowstone at the lake, running from Demersville, most of the way between flat, low
banks; the Big Fork or Swan River, a rocky stream, whose course from Swan Lake to
the Flathead is an oft-repeated alternation of wild rapids and comparatively quiet
reaches; and Dayton Creek on the west, which 1 did not see. The outlet (Flathead
River) flows rapidly away from the lake between bluffy banks which presently become
a canon.

Although this lake lies in a great trough-like valley, the level of much of which is
not far above that of the lake itself, there is scarcely any swampy ground in its vicin-
ity, or weedy standing water connected immediately with it or with its tributaries
in the vicinity of the lake. The principal breeding-grounds offish, in fact, appear to
be upon these streams at a considerable distance from Flathead Lake, so that for most
of the species there is a long migration period.

Our systematic work in the lake was all done in and about the northeast bay in
the vicinity of the mouth of the Big Fork, and at the lower end near the outlet.

While on this bay we were the guests of Mr. K. L. Harwood, of Demersville, and
of the Helena Rod and Gun Club, whose club-house on the bay was our home, while
a steam launch belonging to members of this club afforded the only possible means
of access with our apparatus to the deeper waters of the lake.

At this locality, where we remained from the 20th to the 22d of September, two
dredgings were made, the first beginning at 70 feet and continuing to 125 feet, and
the second beginning at 125 feet and continuing to 153 feet. The surface net was
hauled from 8 a. m. to 9 p. in., in deep and shallow water, and collections were made
with nets and by hand alongshore, among the weeds, from driftwood, and from stones.

Our only temperature observations were made at noon of a bright day (September
22), with a common thermometer only, as no deep-sea thermometer was furnished for
this trip. At this time the temperature of the air was 70° F., that of the water at
the surface 68°, and that of the mud brought up in the dredge, in a haul commencing
at 125 feet and stopping at 153 feet, was 42°.

At the lower end of the lake a heavy storm made work difficult, but we searched
thoroughly a rocky flat at the outlet, and collected from the masses of weeds washed
up by the waves and from the weedy shallows along the southeast shore.

The open-water collections in Flathead Lake were very similar in general charac-
ter and in the relative numbers of the principal groups to those in Yellowstone Lake,
but the species were all different. In the former lake the so called Daphnia yulcx
was not once seen, but this species was replaced by A Duphnin allied to hyalina, and
here described as thorata. This eutomostracau made probably four-fifths to nine-

238 BULLETIN OF THE UNITED STATES FISH COMMISSION.

tentbs of the product of every deep-water haul with the surface net. Diaptomus, the
text ••ommouest form in Yellowstone Lake, was not certainly seen at all in Flatliead,
but was replaced by a new variety of Episvhura (E. nevadensis, var. Columbia-), which
held practically the same relation to Daphnia.thorata which I), sicilis held to D. pulcx
in the other lake. Besides these most abundant pelagic forms we found only occasional
examples of Lcptodora, Cyclops,* Bosmina, Soafholeberis, and Sida cryntallina, the
last two shore forms which probably would- not have been taken very far out. Be-
tween the deeper waters and the weedy northern margin of the northeast bay is an
extensive flat of sand, under from 5 to 15 feet of water, and here our tow-net hauls
were always remarkably unproductive. Partly, perhaps, because of the barrier offered
by this barren belt of shallow water, the pelagic Crustacea did not appear at all in
our alongshore collections as they did in Yellowstone Lake. The assemblage of forms
brought out by the small amount of work inshore which we had time to do, was in
no way remarkable, unless for its deficiencies. Gammarus and Allorchentes dentata
among the amphipod crustaceans, Sida, Eurycercus, and Cyclops gyrln-u* among the
entomostraca, species of Physa, LimiHta, and Planorbis among mollusks, and the usual
miscellany of hydrachnids, epliemerid and Ghironomus larvae, larvse and adults of
Dytiscidfc and llydrophilida;, Corisa, planarians, leeches, and annelids — among the
latter, Pristina lacustris — were the commoner kinds.

Our first dredging in Flatliead Lake was made about 200 yards from land, off the
month of a small cove with bluffy shores — the first below the Helena Club House —
in water ranging from 76 to 125 feet. The dredge came up, after a haul of .about a
quarter of a mile, well filled with soft mud, mostly of slaty color, but somewhat
streaked with reddish brown and mixed with a considerable debris of particles of
dead wood, fragments of dead leaves, cast skins of insect larvae, and the like.

The greater part of the zoological product of this haul was a mass of the cusne
ciuin of a species of polyzoan (Plumatetta, near arethusa), and with these came Chi-
ronomus larvte, red and pale, a dozen specimens of Pixidium, a few cyprids, and a
number of undetermined, slender, pale-red, annelid worms, 2 to 3 inches long and a
millimeter in diameter.

The second dredging was made in the same vicinity, but a little below the pre-
ceding and farther out. Beginning about half a mile out from the head of the same
cove, at a depth of 125 feet, we hauled nearly a mile to south and west, taking up the
dredge at a depth of 153 feet, when about three-fourths of a mile from the point form-
ing the southern limit of the cove. This haul yielded precisely the same product as
the other — an abundance of the same species in approximately the same ratios.

Neither in variety nor quantity was the animal life of the deeper waters of this
lake, as shown by our work with the dredge and towing net, at all in advance of Yel
lowstoue Lake, with the single exception of the polyzoan of our dredgings, and this
was possibly only a local accident.

The bottom and margins of the southern end of the lake seemed comparatively
barren, the weeds washed ashore containing, in fact, scarcely anything but Allor-
chesten dentata, dytiscid and epliemerid larvae, and Corisa. From the stony flat at
the outlet a considerable supply of caseworms of several species was obtained, Corisa,

* Mostly a species undescribed, allied to thomasi of the Great Lakes, but differing in its more
slender, more loosely articulated form and in the armavure of some of its legs.

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 239

water beetles (Hydrophilidce and Dytiscidce), and perlid and ephemerid larvae, together
with a number of leeches — both Clcpsine and Ncphelis — Physa, Limncca, Planorbis,
Pisidium, Gammarim, Plumatclla, and a fresh-water sponge (Spongilla fragilis Leidy).

Swan Lake.— This lake, visited August 24, lies in the course of the river of the
same name, about 8 miles from our camp ou Flathead Lake. It is a long and narrow
lake— about 12 miles long by 1 to 3 miles wide, according to our guide — lying deeply
secluded between two lofty mountain ranges, Mission Kange on the west and the Koo
tenai on the east. Its waters are clear and its bottom is extremely irregular, if one
may judge from the soundings made by us at the lower end, for some 2 miles above
the outlet. The deepest water (not exceeding 30 feet) lay off the highest points, where
the mountains come down to the water's edge, or in pot-holes and winding irregular
channels, with weedy bars and banks between. At the upper end, the lake is said to
expaud to its greatest width and to be surrounded by meadows and marshy flats with
water weeds extending a long distance out; and there are similar weedy flats along
the shore below, especially at the mouths of creeks. Swan River, above the lake, was
also reported to be marshy for some distance up, as is also Spring Creek, emptying
near the head of the lake. About half a mile above the outlet were two small islands
with gravel beaches and surrounded by shoal water full of rushes and a good growth
of other aquatic plants.

The animal species in the deep open water of this lake were precisely the same
as those commonest in the larger lake below; namely, Daphnia thorata, Epischura
ncvadensis, var. Columbia;, taken in numbers, and Leptodora hyalina, occurring only
occasionally. These pelagic entomostracans were, however, much less abundant in
Swau Lake than in the larger, deeper body of water.

The bottom forms were not collected by us, as we had brought no dredge, but the
surface net was hauled repeatedly among the weeds in water about 10 feet in depth, and
stones, round and small, around the margins of the islands were thoroughly searched.
Among the weeds, the commonest eutomostraca were Sida crystallina, Eurycercus
lamellatus, and Cyclops gyrinus — the latter numerous — and with these occurred a very
few specimens of D. thorata and of a species of Alona undetermined. A few small
Chironomus and ephemerid larvae, Gorisa, Agrion larvae, Allorchestes dentata, Physa,
Planorbis, Hydrachnidce, several bristled annelid worms, and a small leech, were also
taken here. From the stones along the margin of the island we took great numbers
of Gammarus and Cypris, a very flue sponge occurring in thickish masses on the rocks
(some white and some chlorophyl-green), a branching polyzoan (Plumatella) clinging
to the stones, several species of leeches, — including Nephelis maculata, so common in
the Park, — planarians, specimens of Physa, Planorbis, and Pisidium, and the usual
aquatic insect forms, larval and adult; viz, larval Ephemeridce and Chironomus and
other small dipterous larvae, Gorisa, aquatic Goleoptera, Hydrophilida;, Dytiscida;, and
several kinds of caseworms. These alongshore collections were, in fact, decidedly
larger and more varied than those from Flathead Lake, a fact doubtless to be ex-
plained in part by the relatively greater amount of shallow weedy water, and the
consequent greater abundance of minute plant and animal life.

240 BULLETIN OF THE UNITED STATES FISH COMMISSION.

DESCRIPTIONS OF SPECIES AND VARIETIES.

CLADOCERA.

Daphnia clathrata, n. s]>.

A species of moderate size, with short, deep head, medium to very long posterior
spine, minute pigment speck, and pectinate tarsal claw. In the immature female there
is a prominent angle just above the swimming antennae, like that of D. dentifcra.

In the adult female the head, measured vertically across the rostrum, is twice as
deep as its length from the base of the antenna to the middle of the front. It is
sharply keeled rather than crested, very broadly rounded, its lower margin very
slightly convex or quite straight, and its rostrum well marked in the adult.

The eye is close to the front, the transparent orbit reaching to the margin of the
head, of medium dimensions, its autero-posterior diameter contained twice in the
space between the eye and the posterior margin of the head. The pigment speck
is very minute, placed behind the lower half of the eye and nearer the posterior
margin of the head. The foruices are not prominent. Beginning midway between the
antenna and the eye, they arch broadly above. the base of the former, making an
obtuse angle a little beyond the antenna, and continuing as a slight carina backwards
and downwards for a little distance on the side of the valve. The ventral margin of
the shell is more broadly arched than the dorsal, the latter being, in the immature
female, nearly straight from the heart backwards. The valves are conspicuously
quadrangularly reticulate, spinose on their lower edges nearly to the beak, and on
the upper edge to the vicinity of the heart. The posterior spine is very long, straight,
slender, spinose to the tip, contained in average cases not more than twice in the
length of head and body without the spine.

The antenna; are rather short, about half as long as the distance from the poste-
rior margin of the eye to the base of the posterior spine. The swimming hairs are
two-jointed, the basal joint the shorter. The dorsal abdominal processes arise in
immediate connection, but are not united at their base. The anal furrow has about
a dozen teeth on each side, and the caudal claw has a comb of three or four conspicu-
ous teeth at its base, besides a little group of smaller ones.

Length of an ovigerous female, 1.7 millimeters to the base of the spine; the greatest
depth, 0.85.

The male was not seen.

Occasional in Grebe Lake, Yellowstone Park.

Daphnia arcuata, n. sp.

Head helmeted, rounded in front, length one third that of the shell, front con-
cave, beak produced, extending beyond the sensory hairs of the antennae. For-
nices beginning above the eye and extending nearly to the middle of the back, not
especially produced above the antenna?. Eye small, about midway between the man-
dibles and the front of the head, and about midway between the tip of the beak and
the dorsal surface of the head. Pigment speck very small, less than half the diam-
eter of a lens of the eye, and placed midway between the eye and the posterior margin
of the head. The latter concave, the buak extending backward and applied against
the margin of the shell. Swimming auteunas reacliing the middle of the shell, their

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 241

hairs moderately robust, the first segment not longer than the second. No emargi-
nation separating head from body, but the dorsal surface very slightly sinuate there.
Abdominal processes all distinct, anterior much the longest. Uaudal claw with
accessory teeth, about six in number, the, three proximal the largest. Anal spines
about ten. Posterior spine of the shell given off from the middle of the valves, in
the adult female a third to a half as long as the valve. Shell moderately coarsely
marked in quadrangular areas, the lower and dorsal margins spiuulose from the
middle backwards. Length 2 millimeters, depth 1 millimeter.
Heart Lake, Yellowstone Park, Wyoming.

Dapbnia thorata, n. sp.

This species belongs to the hyalina group, and may possibly be entitled only to
varietal rank. The distinctness and constancy of its characters, however, in collec-
tions made by us in Flathead and Swan lakes in western Montana, and the number
of minor points in which it differs from hyalina, as most recently described, lead
me to distinguish it here as a specific form.

It is oval in outline; the long and slender posterior spine is placed at or a trifle
above the middle; the length of the head is about a third that of the valves of the
shell excluding the spine, and there is no trace of dorsal emargination between head
and body. The head is narrowed toward the base and elongated forward in a way
to give it the outline of a high bell jar with a flaring base. Its front is broadly
and regularly rounded, its ventral margin usually conspicuously concave and closely
like the dorsal, although occasionally the head is straight or convex beneath. The
posterior margin is either straight or slightly concave, and the beak stands free from
the front margin of the valves, and by its extension downward not only covers the an-
ten use but reaches clearly beyond the tips of the sensory hairs. The eye is of medium
size, placed far back of the front of the head and equidistant from the tip of the beak
and the dorsal junction of the head an d body. The pigment speck is of moderate size,
placed directly behind the eye, and much farther from it than from the posterior mar-
gin of the head.

The antenna} are moderately stout, entirely smooth except for inconspicuous trans-
verse rows of minute appressed hairs upon both peduncle and rami, and a row of
short, tooth-like spinules at the distal end of each segment. The swimming hairs are
rather slender, the second joint commonly decidedly shorter than the first.

Pornices slight, arising above and a little behind the eye and terminating directly
behind the antennae, above the bases of which they project but slightly. The lower
margin of each valve is set with the usual spiuules almost to the beak, and the dor-
sal margin is similarly armed for a distance in front of the spine about equal to half
the length of the latter. The valves are marked off by fine lines into large quadrate
meshes.

The dorsal abdominal processes rise separately, the two anterior, however, in
immediate contact at their bases. . The first of these is decidedly the longer, but the
third process is distinct, although low. The anal setae are two-jointed, the second
joint the shorter. The abdomen is regularly narrowed backwards, and the anal
groove is provided with twelve to fifteen teeth on each side, commonly the" latter
number. The terminal claws are without accessory comb. The intestinal cceca are
short, not longer than the diameter of the eye, and extend directly forward.
F. C. B. 1891—16

242 BULLETIN OF THE UNITED STATES FISH COMMISSION.

Length, 2 millimeters to '2.5 millimeters; depth half the length, sometimes a
little more. Length of spine somewhat variable, but commonly about equal to the
depth of the sheik

Described from females only.

Abundant in Swan and Flathead lakes, Montana.

Daphnia pulex, var. pulicaria, n. var. (Plate xxxvn, Fig. 1.)

Similar, especially in the female, to typical D. pulex, to which it is closely re-
lated by its more general characters. Body a broad oval, moderately thick, colorless,
commonly without dorsal emargination between the head and thorax, although some-
times in the generation of females bearing ephippia there is a broad concavity just
above the heart. The lower border of the head is broadly concave and the beak is
long and applied against the anterior margin of the shell. Moderately long posterior
spine placed above the middle line; caudal claws with two sets of teeth, and with 14
to 1 7 curved spines at the anal furrow.

The head of the female is small, somewhat depressed, crested, as in D. pulex, the
crest extending backward to the middle of the dorsnm; fornices terminating pos-
teriorly opposite the heart, and extending anteriorly to the eye. The beak projects
a little beyond the tips of the sensory hairs; the eye is large, its vertical diameter
contained scarcely twice in the distance from the eye to the tip of the beak, placed
close to the broadly rounded anterior margin of the head, and provided with many
large lenses. Pigment speck of moderate size, midway between the eye and the
posterior margin of the head.

Antenme but moderately developed, destitute of scale-like appendages like those
ofpule.r, but set with inconspicuous transverse rows of rather slender hairs. Swim-
ming hairs moderate and moderately feathered, three-jointed, the third segment
very short, but evident. The coecum of the intestine strongly curved, extending at
first obliquely downwards towards the middle of the eye, and then turning almost
directly upward at an acute angle, terminating midway between the middle of the
upper margin of the eye and the front of the base of the antenna. The surface of the
valve is marked with quadrate areolations, and the margins, both dorsal and ventral,
are provided with backward-projecting spines or thorns as far as the middle. The
anterior half of both margins smooth. The posterior spine is variable in length,
reaching in adult females a fourth the entire length of the head and body.

Dorsal processes of the abdomen distinct, the two anterior contiguous in their
origin, not united at their base, the first the longer, smooth, and directed forward,
the second hairy, turning backward. Two others -in the form of low elevations, the
last inconspicuous, but both hairy.

The abdomen is rather broad, the posterior margin broadly rounded, the ante-
rior margin straight; 13 to 17 spines bordering the anal furrow, length regularly
increasing from above downwards; the teeth of the caudal claw in two groups of
from four to six each, the upper group very much the smaller; the anterior margin of
each claw with two distant slight emarginations, as in D. pulex.

Length of an adult female 1.9 millimeters without the spine; depth, 1.1 millimeters;
spine 0.5 millimeter. Female bearing ephippium, a little deeper (1.2 millimeters).

The male smaller, narrower, with head more depressed, the dorsum especially more
nearly straight, and the posterior spine standing higher, continuing the line of the
dorsum backwards. The lower margin of head is only slightly concave, the posterior

AQUATIC INVERTEBRATE FAUNA OP WYOMING AND MONTANA. 243

half of it straight. A slight beak is formed just below the sensory antennae, the latter
being attached at a small angular einargiuatiou at the posterior angle of the head.
From this emargination the posterior margin of the head passes directly upwards in
a broad and gentle curve. The eye is very large, placed at the very front of the head.
Its longitudinal diameter is contained but once in the head behind the eye. Sensory
antenna slightly clavate, slightly expanded at the middle, its length equal to the
vertical diameter of the eye. In front of the terminal group of sensory hairs is a long
termiual spine, nearly as long as the antenna itself, slightly curved backwards and seg-
mented at the middle. Accessory hair distant from end, but a little below the middle.

Length without spine, 1.4 millimeters; depth 0.9 millimeter; spine, 0.33 millimeter.
A single hairy dorsal abdominal process, as in pulex.

Yellowstone Lake and other waters of Yellowstone Park.

Daphnia dentifera, n. sp. (Plate xxxvn, Fig. 2.)

This species is broad oval in form, has a long beak and a very large eye, a poste-
rior spine placed high up, and in the male and young female a prominent angle on the
dorsal outline between heart and eye.

The head is broadly rounded, with eye close to the front margin. The foruices are
short, rising above and behind the eye and extending backwards a little beyond the
base of the antenme, where they form a prominent angle. Thence a slight lateral keel
of the valve is continued downwards and backwards a distance about equal to the
length of the fornix. The lower margin of the head is broadly concave, the beak pro-
duced, projecting as far as the ends of the sensory hairs. The large eye, with numer-
ous lenses, is contained not more than twice in the distance from eye to beak, its
diameter a little greater than that of the base of the antenna at its insertion. Pigment
speck of moderate size, circular, immediately behind the eye and nearer to that than
to the posterior margin of the head.

The head is slightly crested, and the crest, extending backward to the heart,
rises over the antenna? in a prominent, nearly rectangular process, still more acute
in the young, the tip of which is commonly truncate and bears two or three teeth
inclining forward. In the egg-bearing female this process is reduced to a mere
obtuse angle, or, in the last generation (that bearing the ephippinm), disappears
entirely. In young adults this dorsal angle is midway between the eye and the heart,
but when fully developed it is on a line drawn from the anterior margin of the valve
to the middle of the base of the antenna. The seta} of the antennae are all two-jointed,
the basal joint distinctly the longest. The posterior spine of the carapace is long,
slender, and weak, and is commonly contained three or four times in the head and body
without the spine.

The margins of the valves are set below and behind with slender thorns, as is
also the posterior spine, these thorns extending forward a little distance upon the
dorsal margin of the shell. The curved spines bordering the anal furrow are thirteen
in number; the caudal claws are without accessory teeth; the surface of the shell is
marked with quadrangular areolations.

The first and second dorsal abdominal processes are about equal in length and
arise in immediate contact, the anterior turning forward and the posterior backward.

Mature female 1.8 millimeters long by 1 millimeter deep.

The male of this species is smaller than the female; the head is smaller and nar-
rower, the form is more nearly elliptical, and the dorsal angle is as prominent as in

244 BULLETIN OF THE UNITED STATES FISH COMMISSION.

the young female, and commonly bears obscure teeth at the tip. The head is sub-
quadrate, with rounded angles. The very large eye is at the extreme front of the head,
its diameter greater than the distance between the eye and the posterior margin of the
head. Below, the head is straight; the anterior antennae are not especially prom-
inent and the terminal spine is inconspicuous. The posterior spine is like that of the
female, long and slender and dorsally placed. The abdomen is without dorsal process.
Mature specimen 1 millimeter long by 0.5 millimeter deep.

Closely allied to D. dentata Matile, with which iny friend Professor Birge con-
siders it possibly identical. It differs, however, particularly in the form of the head,
the beak of which is much more produced backward in dentifera than in dentata;
in the somewhat larger eye (especially of the male) ; in the different form and posi-
tion of the dorsal angle, and in its evanescent character in the female adult; in the
greater length and sleuderness of the posterior spine ; and, notably in the male, in the
different armature of the anterior antenna. Dentifera is also without the tliird joint
of the swimming hairs of the antenna.

Pool near Shoshone Lake, Yellowstone Park.

OSTRA.CODA.

Cypris barbatua, n. sp. (Plate xxxvn, Figs. 2 and 3, and Plato xxxvni.)

An extremely large, very hairy, oblong Cypris, with rounded ends and dorsal
and ventral margins nearly parallel. Length, 4 millimeters; width, 1.6 millimeters;
depth, 2 millimeters. A very little deepest at hind end of hinge margin. (Depth
across eye, 95 per cent of greatest depth.)

Dorsal margin about straight for a great part of its length, the ventral margin
very slightly emarginate or sinuate at its anterior third. The anterior end broadly
and smoothly rounded, more obliquely above than below, the posterior somewhat
obliquely rounded, the ventral margin being thus nearly half as long again as the
dorsal. Seen from above the shape is symmetrical, a slender oval, a little more
flattened at the sides behind than before; thickest, consequently, before the middle.

Color a dirty yellowish-brown in alcohol, with a reddish-brown patch on either
side above and behind the middle. Surface of valves opaque, very minutely rough-
ened, and well covered with conspicuous hairs, which give this Cypris a decidedly
hairy appearance to the naked eye. Hairs longest before and behind and length-
ening generally towards the margin, where they project as a fringe, the most prom-
inent part of which is a row of hairs borne on slender conical tubercles within the
margin of the valves. The valves are equal and the shell fairly full, but not plump.

Anterior antenna with the basal segment obliquely channeled, partially dividing
it into two, the distal part of which bears a single bristle on its superior surface,
and two long, more slender ones, springing together from the tip of the ventral sur-
face. A short, subquadrate second segment bears a single seta, about as long as the
segment, on the dorsal surface, near the tip. From the distal end of the following
segment spring two long, slightly plumose setse, one dorsal, one ventral, the former
much the longer. The fourth segment bears at its tip four long setsc, two of which
arise from the ventral angle and two from the outer dorsal. The following segment
is similarly armed, and the distal extremities of the sixth and seventh are densely set
with long plumose set;e forming a stout fascicle, which extends beyond the end of the
antenna a distance equal to the length of the antenna itself.

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 245

The terminal segment of the palp of the first maxilla is a little more than a
fourth the length of the basal, the latter with one subterininal bristle without, and
several terminal ones. Tip of last segment with two stout, curved, claw-like setse,
and four or five smaller, softer ones. Outer lobe of maxilla proper reaching to tip
of first segment of palp, nearly equaling it in diameter, also with two curved claws,
shorter but much stouter than those above mentioned, three-fourths as long as the
lobe itself. Besides these, two smaller setae and three or more subterininal ones, two
of which are smooth, like the terminal group, and one strongly plumose. A single
plumose seta also springs from near the base of the concave surface of this lobe.
The second and third lobes similarly armed at tip, but with a larger number of
curved setae, all of which are soft. Two of these, on the sliort inner lobe, are much
longer and stouter than the others, and project directly backward. The base of this
lobe bears two plumose setee about as long as those just mentioned. The length of
the inner lobe is half that of the outer, the middle one being intermediate.

The second maxilla with about twelve terminal setae, which diminish in length
inward, most of them slightly plumose, and two long slender setae, one springing from
the middle of the inner margin and the other from the base. Palp thick, slender
ovate, twice as long as the masticatory lobe, fringed with a soft silky pile, and bearing
three more or less plumose setae at its tip, the middle one of which is the longest.
Branchial lobe very small, semicircular, with three fully developed plumose setae
nearly as long as the palp, and two much shorter ones, one delicate and smooth, the
other stout and plumose.

The basal segment of the second antenna trigonal, with one moderately long hair
beneath, and two of similar length springing together from the inner side of the apex.
The.secoud segment subcylindrical, with two hairs diverging from the middle of the
outer side of the apex, the under one of which is very short and weak, about as long
as the third segment is wide, while its companion reaches about to the tip of that
segment. On the inside of the tip of the second segment is another hair, similar to
the above, and of about the same length. The third segment bears, at the union of
its basal with its middle third, on the under side, set beyond a slight tooth-like pro-
jection, a jointed olfactory club, whose length is about two-thirds the diameter of the
segment. Otherwise this segment bears no hairs except at the tip, where, upon its
inferior angle, is one long, stout hair, reaching beyond the tip of the last joint, and
upon its inner surface a fascicle of five plumose hairs, the four longer of which are
curved and parallel, while the fifth is short and straight. The third segment is slightly
longer than the second and about two-thirds as thick. The fourth segment is three-
fourths the length of the third and about two-thirds its diameter, slightly enlarged
at the middle, where it bears, on the under side, a group of three long hairs, and upon
the upper side two shorter ones. At the tip of this segment are a group of three
long plumose hairs and a stout, curved, concave, acute claw,, nearly three times the
length of the last segment, doubly dentate on both edges. At tip of last segment the
usual strong, curved, bidentate claws, five in number, three of equal length, as long
as the two last segments of the antenna, and two others about half that length.

Mandible with a row of six dark corneous teeth, more or less bifid, the series
continued in an irregular cluster of tooth-like spines, and terminating in two highly
plumose setae. The series of teeth with numerous accessory smaller teeth and spines,
and two transparent lamellae — slender, but as long as the teeth themselves — inserted

246 BULLETIN OF THE UNITED STATES FISH COMMISSION.

between the first and second and the second and third series, respectively. The latter
lamella is recurved and serrate on its concave edge. Basal segment of the palp long-
est, the third next, second and fourth subequal in length, the second as broad as the
first. The latter bears at its posterior tip three plumose setae of unequal length, in
a cluster, and a fourth larger, stouter, decurved articulate one, inserted on the outer
side of the tip of the segment. The second segment has in front a group of three
slender setae inserted a little behind the tip; and opposite to them upon a stout tuber-
osity another group of three long equal sette, to which a fourth stands in the same
relation as on the preceding segment. On the third segment is a group of five setae
similar to those on the anterior margin of the segment preceding, and, in addition,
a circlet of six, attached around the posterior and inner margin of the end of this
segment. At the tip of the palp are three curved claws, averaging as long as the
two preceding segments together, with some slender setae intermixed. The so-called
branchial appendage is about as long as the basal segment of the palp, and bears four
stout plumose setae with a small accessory seta hi front.

First leg with basal segment columnar, distal portion partially separated, without
hair or bristle. Second segment cylindrical, its surfaces smooth except for numerous
transverse rows of exceedingly fine short setie, present also on the two succeeding
segments of this leg. A stout bristle at anterior distal angle. Third and fourth seg-
ments nearly equal (the third, however, somewhat the longer), together slightly longer
than the second, the length of each about twice its transverse diameter. The third
with a single apical hair at the anterior angle, and the fourth with but two, one of
which, is as long as the segment itself, and the second about half that length. Termi-
nal segment with a very long, slender, symmetrically curved, regularly tapering claw,
with two short soft setae springing from its base. The entire claw somewhat longer
than the last three segments conjointly.

Caudal raini long and slender, slightly sinuate, the transverse diameter of each not
more than one-twentieth its length; the basal fifth, however, considerably thickened.
Kami smooth, except posteriorly, where the margin is closely set with stout, short
spines, lengthening toward the distal end of the ramus. Terminal claw slightly
curved at tip, contained two and a half times in the length of its ramus. Subterininal
claw nearly two-thirds the length of the terminal, also slightly curved. Claw-like seta
almost immediately above the latter, more slender, but two-thirds its length. Besides
the above, a short slender seta springs from in front of the base of the terminal claw.

The first and last segments of the second pair of legs subeqnal, each two-thirds
the length of the second; basal segment straight, its length five times its width, with
three slender seta}, one borne upon the middle of its exterior side, and two near the
apex, opposite each other. Second segment slightly curved, with a single slender seta
near the apex, on its outer margin. Third segment with two terminal setae, one nearly
straight, and claw-like, about three-fourths the length of its segment, and the other
curved and blunt.

This species may be the same as C. grandis Chambers*, which it certainly seems to
resemble closely, but from which it differs, if I may judge from the published brief
description and rude figures, in color, surface, form, arrangement of auteunal setoa, and
other minor details.

Yellowstone Eiver, Yellowstone Park, Wyoming.

*" New Entomostraca from Colorado," in Bull. TT.S. Gool. anil Geogr. Surv., vol. in, No. 1, p. 151.

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 247

COPEPODA.

Cyclops minnilus, n. sp.

A small slender species, with seventeen-jointed antennae, with narrow and loosely
articulated ceplialothorax and salient thoracic angles, slender abdomen, long and
narrow furca, and but two well-developed caudal seta; for each ramus. The antennae
reach to the posterior margin of the second distinct segment, and are of very nearly
the length of the abdomen (including furca, but excluding the caudal setee). The
greatest width of the thorax is contained two and one-third times in its length, and
the furca is very nearly half the length of the remainder of the abdomen. The diam-
eter of a ramus is about one-seventh its length.

The rudimentary inner caudal seta is a trifle longer than the outer, and about
a third the length of the ramus; the longest seta as long as abdomen and furca; the
next in length less than half the longest.

The last segments of the thoracic legs are armed as follows:

First pair: outer ramus, one spine and two setae at tip, two setae within, and one
seta without; inner ramus, one spine and one seta at tip, three setae within, and one
seta without.

Second pair: outer ramus, one spine and one seta at tip, three setae within, and
two spines without; inner ramus, one spine and one seta at tip, three setae within, and
one seta without.

Third pair: outer ramus, two spines and one seta at tip (second spine twice as
long as first), three setae within, and one spine without; inner ramus, one spine and one
seta at tip, three setae within, and one seta without.

Fourth pair: outer ramus, two spines and one seta at tip (second spine twice as
long as first), three setae within, and one spine without; inner ramus, two spines at
tip (one twice as long as the other), two setae within, and one seta without.

Rudimentary legs of fifth pair distinctly articulated, basal article with a long seta
at its outer distal angle, and second article with two seta? at its blunt tip, the outer
the Jonger.

Duck Lake.

Cyclops serratus, n. sp.

A very long, narrow, loosely articulated species, with strikingly salient thoracic
angles ; cephalothorax broadest far forward and lobed in front, between the seventeen-
jointed antennae.

Abdomen long and slender, with very long and narrow caudal rami, and but two
developed caudal set* to each ramus. The first segment is but little longer than wide
(eight to seven), is broadest across the middle, and excavate in front at the base of
each antenna, leaving a thick, median, projecting lobe. The second segment is nearly
a fourth as long as the first, and but little narrower, broadest across its posterior
angles, which, though blunt, are so strongly salient that the lateral margins are
decidedly sinuate. The third segment is as long as the second, but narrower, and
with its sides more nearly parallel. The fourth and fifth segments are progressively
shorter and narrower, the latter being trapezoidal, as seen from above, and separated
from the first abdominal segment by a deep acute emargination.

248 BULLETIN OF THE UNITED STATES PISH COMMISSION.

The abdominal segments are as long as the cephalothoracic segments two to five
taken together, and the fnrca is as long as the last three segments. The first segment
of the abdomen 'is broadest in front, where its width is nearly as great as its length.
The second is as broad as long, the third and fourth equal, the fifth a little shorter,
the last with a row of fine spiiiules around the base of the ranii.

The width of each rainus is contained nearly eight times in its length. Besides
the lateral spine — situated a little before the posterior third of the ramus — there is a
cluster of two or three minute spines at its anterior fourth. The outer and inner
terminal setae are reduced to short subequal spines about twice as long as the ramus
is wide. The other setae are slender, plumose, the inner nearly twice as long as the
outer of each pair.

The antennae are rather stout and short, seventeen-jointed, reaching to the end
of the second segment. They are without special structures or appendages.

The last segments of the thoracic legs aie armed as follows:

First pair : outer ramus, one spine and two setie at tip, two setre within, and one
spine without; inner ramus, one spine and one seta at tip, three setae within, and one
seta without.

Second pair: outer ramus, one spine and one seta at tip, three setce within, and
two spines without; inner ramus, one spine and one seta at tip, three seca3 within, and
one seta without.

Third pair: outer ramus, two spines and one seta at tip (one spine twice as long
as the other), three setae within, and one spine without; inner ramus, one spine and
one seta at tip, three setie within, and one seta without.

Fourth pair: outer ramus, two spines and one seta at tip (one spine double the
length of the other), three setae within, and one spine without; inner ramus, two spines
at tip (one double the length of the other), two setae within, and one seta without.

The fifth pair is two-jointed, the basal joint broad, Quadrate, with a seta at its
outer angle; the second cylindrical, with one long and one short seta at tip.

Length, without setae, 1.34 millimeters.

Described from females only.

Cyclops capilliferus, n. sp. (Plate XL, Figs. 14-17, and Plate XLI, Fig. 18.)

This is a symmetrical, compact Cyclops, with the cephalothorax closely articulated,
widest at the middle, and the sides regularly convex, with the abdomen narrow and
slender, with three well-developed caudal setie, and sixteen-jointed antenna bearing
several very long setie.

The abdomen, with caudal furca, is contained a little less than twice in the ceph-
alothorax, and the breadth of the latter is just half its length. First segment very
long, five times the length of the second ; second and third equal ; the fourth very
short, on the median line semicircularly excavate behind. The abdominal segments
in the female diminish regularly in length from first to last. The caudal rarni are
twice the length of the last segment and one-fourth as broad as long. The lateral seta
is placed a trifle beyond the middle of the ramus; the outer terminal seta is a short
naked spine; the other three are well developed and plumose. The inner and outer
of these three are of nearly equal length, the latter a little the longer, the middle one
much the longest one of the group.

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 249

Antenna moderate, reaching about to the middle of the second segment of the
cephalothorax. Sixteen-jointed in all adult females, and further especially distin-
guished by the presence of very long flexible setae upon the first, third, tenth, and
fourteenth segments. Terminal setae likewise very long. The seta borne by the first
segment extends to the twelfth ; that of the third reaches to the fourteenth ; that upon
the tenth segment extends to the tip of the antenna, and that upon the fourteenth far
beyond it. All these foregoing setae are borne upon the anterior terminal angles of
their segments with the exception of that of the fourteenth, which is borne upon the
posterior angle.

The first segment is as long as the two following, and very nearly twice as long
as wide. The second is very short, its length one fourth its width, and the length and
width of the third are equal. Of the three terminal segments the penultimate is
longest, being twice as long as broad ; the antepenult two-thirds the length of the
following; the last is about as wide as long.

The last segments of the thoracic legs are armed as follows :

First pair: outer ramus, one spine and two setae at tip, two spines without, and
three setna within; inner ramus, two setae at tip, one within, and two without.

Second pair: outer ramus, one spine and two setae at tip (the inner of the latter
slender, the outer thick), four setae within, and two spines without; inner ramus, two
setae at tip, one within, and three without.

Third pair: outer ramus, two setae at tip (the outer one stout, short, and spine-
like), four setae within, and two spines without; inner ramus, two setae at tip, three
within, and one without.

Fourth pair: outer ramus, two setae at tip, four setae within, two spines without;
inner ramus, two setae at tip, two within, and one without.

The fifth pair are two-jointed, the terminal joint with one long and one short seta
at tip; the basal joint with one long seta without.

Length, without setae, 1.2 millimeters.

Grebe Lake, Yellowstone Park.

Cyclops thomasi Forbes. (Plate xxxix, and Plate XL, Fig. 13.)

Cyclops thomasi Forbes, Amer. Nat., xvi, Aug. (1882), p. 649; Cragin, Trans. Kan. Acad. Sci.,
vin, 1881-82, p. 68 (1883); Herrick, Final Report, p. 153 (1884); Underwood, Bull. 111.
State Lab. Nat. Hist., u, 1886, p. 332; Forbes, Rept. U. S. F. C., 1887, p. 707 (1891).

A long and slender species, with seventeen-jointed antennae, oval cephalothorax,
somewhat closely articulated, slender abdomen, very long and slender caudal rami,
and two developed setae to each ramus, the longer of which is about twice as long as
the shorter.

The cephalothorax is widest at about the middle, its greatest width a little more
than half its length. Posterior angles not prominent or produced, except those of
the last segment, which are slightly produced outwards. Sides of the first segment
subparallel, rounding slightly towards the front, the segment itself twice as long as
the other segments combined; the second segment shorter than the third, but longer
than the fourth; the fifth reduced to a narrow linear band, as seen from above, the
extremities of which project a little beyond the lateral outline.

Abdomen, with furca, a little shorter than the cephalothorax, its greatest width
one-fourth of its length, including furca. First segment in the female as long as all

250 BULLETIN OF THE UNITED STATES FISH COMMISSION.

the others together, broadest in front, its lateral outlines emarginate behind the
anterior angle. Posterior margin of last segment serrate beneath and at sides; those
of other abdominal segments smooth. Furca as long as the last three segments, the
width of the rami about one-seventh of their length. The inner of the two longer
setae as long as the entire abdomen, the outer of the two half that length. The ex-
treme outermost of the terminal sette two-thirds the length of the inner; that is, about
one-fourth the length of the caudal ramus. Kami slightly curved outwards, with one
large spine and a few small ones a little beyond the middle of the outer surface, and a
vertical comb of small spines at one-fourth the distance from the proximal end.

Antennae of the female moderately robust, reaching about to the middle of the
third segment, without special accessory structures or appendages, the three terminal
segments gradually increasing in length, the antepenultimate two fifths the length of
the last. The two segments preceding the former, taken together, shorter than the
last segment and about equaling the penultimate.

First pair of legs : outer ramus, two setae at tip, two spines without, and two setae
within; inner ramus, one spine and one seta at tip, one seta without, and three setae
within.

Second pair of legs : outer ramus, one spine and one seta at tip, two spines with-
out, three setae within; inner ramus, one spine and one seta at tip, one seta without,
and three setae within.

Third pair of legs : outer ramus, one spine and one seta at tip, two spines without,
and three setae within; inner ramus, one spine and one seta at tip, one seta without,
and three spines within.

Fourth pair of legs : outer ramus, one spine and one seta at tip, two spines with-
out, and three setae within ; inner ramus, two spines at tip, one seta without, and two
setae within.

The outer ramus of the first leg is so foreshortened that the distal outer seta seems
to be placed at the tip of the segment, but the usual tooth marking the lateral distal
angle of the segment stands between this point and the seta next within, thus show-
ing that the spine should be counted as lateral.

The terminal spines of the inner ramus of the fourth pair are unequal, the inner
one a little more than half the outer.

Fifth pair of legs of two segments, the basal segment about as long as broad, with
a strong plumose spine from the outer angle, the terminal segment cylindrical, twice
as long as broad, with two terminal setae, the outer of which is as long as the seta of
the preceding segment, and the inner a little more than half that length.

Total length, without setae, 1.33 millimeters; greatest depth a trifle less than one-
third the length of the cephalothorax.

The common Cyclops of Yellowstone Lake, occurring also in various other waters
of that region. This well-marked and constant species has a range at least from the
Rocky Mountains to the Atlantic region, being, according to Prof. Cragin, a common
species in the water supply of Boston. It is also the usual Cyclops of the Great Lakes.

The original description was inaccurate in two particulars : the outer distal spine
of the outer ramus of the first leg was called terminal, and, by typographical error,
the terminal joint of the outer ramus of the second, third, and fourth legs were said
to have two setae within instead of three.

AQUATIC INVERTEBRATE FAUNA OF WYOMING AND MONTANA. 251

Diaptomus shoshone, n. sp. (Plate XLII. Figs. 23-25.)

A very large and robust species. Thorax broadest in front, across the maxillae,
tapering gradually, with little convexity, to the posterior third. In the female the
angle of the last segment is bifid, both projecting points being minutely spinose at tip.
The first segment of the abdomen is laterally expanded ; the expansion of the left side
with a minute spine at the apex behind; that on the right produced at the same point
into a small, prominent, rounded tubercle, 0.03 millimeter in length, about as broad
as long, making this first segment somewhat unsymmetrical. This is not merely a
modified cuticnlar appendage, but is penetrated by the hypodermic. Egg mass very
large, obovate (narrowest forward).

Right antenna of male robust, the last two joints without special appendages,
antepenultimate with a very long inarticulate process at its outer apex, extending be-
yond the tip of the penultimate, and to the middle of the last segment. The margins
of this process are smooth, but it is broad and emarginate at the tip.

The fifth pair of legs of the male reseinble the corresponding appendages of Diap-
tomus stagnalis, but differ notably in detail. The left ramus of the right leg is borne
at the inner terminal angle of the second joint; is longer than the joint following; is
armed at the apex with a few small acute spines; and bears upon its outer margin,
near the tip, a broad fascicle of delicate hairs. The basal joint of the outer ramus is
two-thirds the length of the second joint of the peduncle, and without hairs or spines
of any description. The second joint of this ramus is about equal in length to the
second joint of the peduncle, and bears at its outer margin, close to the tip, the usual
stout seta, which is two-thirds as long as the joint to which it is attached. The ter-
minal claw is not regularly curved, but is nearly straight for the basal three-fourths.
The left leg is biramose, the inner ramus straight, slender, extending about to the
middle of the second joint of the outer, and armed at its tip. The second joint of this
ramus is as long as the first, if measured from the tip of the apical spine. This spine,
seen from behind, is stout, conical, rather blunt, and has opposed to it withiu, pro-
jecting from the inner angle of the segment, a stout, curved seta, slightly plumose on
it's distal half. Between these, but more closely applied to the outer spine, is a hemi-
spherical cushion-like elevation, set with small, short spinules. On the basal half of
the inner margin of this terminal segment is also a much larger hemispherical cushion,
but with longer and more slender hairs, while the terminal half of the inner margin of
the segment preceding is also moderately inflated and covered with delicate hairs.

The antennae of the female are 25-jointed, as usual, and reach to the base of
the abdomen. The legs of the fifth pair closely resemble those of stagnalis, but
have the terminal setae of the inner rami much less developed. This ramus is a
little shorter than the basal joint of the outer ramus, and of about half its diameter.
It bears at its tip two stout setae equaling the ramus itself in length, plumose under a
high power, and has, in addition, at its inner tip and on the inner margin adjacent, a
patch of delicate hairs and spines. The second joint of the outer ramus is as long as
the first, if measured to the tip erf its terminal claw. The latter is nearly straight,
very slightly recurved. This joint bears a single spine at its outer distal angle, just
within which is the rudiment of the third segment of the ramus, which bears two
spines similar to the above, the inner of which is the longer, the outer itself being
longer than the adjacent spine of the second joint. Adults of both sexes are blood-
red throughout, except the egg sac of the female, which is purple.

252 BULLETIN OF THE UNITED STATES FISH COMMISSION.

Dimensions of female : Length to tip of caudal setae, 3.1 millimeters ; abdomen, with
setae, 1.16 millimeters; without, 0.67 millimeter; thorax, 1.95 millimeters in length;
depth, 0.725 millimeter; width, 1 millimeter.

Male averaging scarcely smaller, but somewhat differently proportioned. Thorax,
1.85 millimeters in length; depth, 0.58 millimeter; width, 0.08 millimeter; abdomen,
without setae,, 0.745; with setae, 1.35 millimeters.

Especially abundant in Shoshone Lake, but occurring in other lakes and even in
pools of some size in Yellowstone Park.

Diaptomus lintoni, n. sp. (Plate XLII, Figs. 26-28.)

A large red species occurring commonly with I), shoshone, but distinguishable
from it at a glance by its different shape, its longer antennae, its smaller size, and by
characters derived from the right antenna and the fifth foot of the male. The thorax
is symmetrically elliptical in shape, broadest at the middle. The posterior angles are
not produced or bifid, but are each armed with a minute spine. The first segment of
the abdomen of the female is not especially produced, but bears at its broadest part a
minute spine on each side. The abdomen itself is very short, its length contained
about three and one-third times in that of the cephalothorax. The antenna of the
female is long and slender, 25-joiuted, reaching a little beyond the tip of the abdomen.

The fifth pair of legs in this sex is similar to those of D. shoshone, but much
smaller. The inner ramus is not jointed. It is longer than the basal joint of the outer
ramus, bears two stout plumose setae at its tip, somewhat shorter than the ramus
itself, and has likewise at its inner tip a patch of small spines or fine hairs. The second
segment of the outer ramus with its terminal claw is two-thirds as long again as the
preceding segment, the breadth of the latter two-thirds its length. The third joint is
indicated by a single long stout seta and one or two smaller ones.

In the male the geniculate antenna is relatively rather slender, its last two joints
without special appendages, its penultimate with a slender transparent apical process,
reaching about to the middle of the succeeding segment, acute at tip, but neither serrate
nor emarginate. Fifth pair of legs in the male usually without internal ramus to the
right leg, but this ramus sometiuies represented by a small rudiment. The liinb is
usually slender and its terminal claw short. The basal segment of the outer ramus
is nearly as long as the adjacent segment of the pedicel, and the slender second
segment of this ramus is fully as long. Long lateral spines borne near the tip of this
segment. The terminal claw is about two-thirds as long as the segment, is somewhat
abruptly angulated near its base and slightly recurved at the tip. The inner ramus
of the left leg is very stout and long, reaching almost to the tip of the outer ramus,
is slightly curved outwards and has the apex minutely hairy. The basal segment of
the outer ramus is thick, two-thirds as broad as long, somewhat inflated within, where
it extends downward beyond the articulation with the second segment as a rounded
expansion covered with extremely fine hairs. Second segment of this ramus longer
than first, but only half as wide, bearing at its tip, within, a rather small, obliquely
projecting cushion covered with cilia, and with two stout terminal spines, one short,
blunt, straight, and smooth, the other curved and plumose, its length about half that
of the segment to which it is attached.

The total length of this species is about 2.5 millimeters, excluding caudal setae;
depth, 0.42 millimeter.

AQUATIC INVEETEBEATE FAUNA OF WYOMING AND MONTANA. 253

This species is closely related to D. stagnalis Forbes, from which it differs con-
spicuously by its smaller size, more symmetrical cephalothorax, without prominent
or bifid angles, and longer and more slender antennae, with longer and more slender
appendage to the antepenultimate segment.

In the fifth legs of the female this species differs from stagnalis especially with
respect to the inner ramus, which is larger and longer than in the other, lacks the
characteristic segmentation of stagnalis, and bears at its tip shorter and broader set:e.
In the male the terminal claw of the outer ramns of the right fifth leg is mnch more
slender than in stagnalis, and the inner ramus is much less developed. The left leg of
this pair is different in a number of details, especially in the length and strength of
the inner ramus and the length and dissimilarity of the set»3 at the end of the outer.

Common in lakes and pools of Yellowstone Park.

Named for my friend, and companion on the trip of 1890, Prof. Edwin Linton, of
Washington and Jefferson College, Pennsylvania.
Diaptomus piscinae, n. sp. (Plate XLI, Fig. 22.)

A species of medium size and symmetrical proportions, antennae reaching to the
tip of the abdomen, cephalothorax broadest about the middle, with fovir distinct
sutures, the posterior lateral angles not produced, but armed with two distal spines.

The right antenna of the male is without appendage to the antepenultimate
joint, and the fifth pair of legs of the same sex has the inner ramus well developed on
both the right and left sides. The usual length is 1.75 millimeters, the transverse
diameter 0.45 millimeter; the abdomen, with furca, is a little more than one- third the
length of the cephalothorax.

The fifth pair of legs of the female is without especially marked characters, except
that the inner ramus, which reaches to the tip of the principal segment of the outer, is
provided with two long, stout, equal setee more than half as long as the ramus itself.
The third joint of the outer ramus is aborted, and bears two short, stout spines, and
the joint preceding bears a slender spine outside the base of the last. The terminal
claw of this joint is simple and nearly straight, viewed in the usual position.

In the male the fifth pair of legs has a considerable resemblance to the corre-
sponding appendages of 1). leptopus, from which, however, this species differs by its
more slender form and by the absence of the autennal hook. The peduncle of the
left leg is quadrate and equal in length to the basal segment of the outer ramus, but is
nearly twice as wide. The sides of this latter segment are parallel, the inner terminal
angle is broadly rounded and minutely ciliate, and to the outer terminal angle is attached
the second segment of the ramus. This segment is a trifle shorter than the preceding
and less than half as wide, and bears at its tip a stout, blunt, conical spine, whose
length is equal to that of the diameter of the ramus, and within this a long flexible
hair as long as the ramus itself. The inner ramus of this leg is very long, reaching
beyond the middle of the terminal joint of the outer ramus. It is slightly concave
towards this ramus and terminates with a broadly rounded or subtruucate, thickly
ciliate end, forming an acute outer angle and an obtuse inner one. Seen at right
angles to this view, the tip is simply obtusely pointed.

The right leg of the male is without remarkable distinguishing characters. Basal
joint of the outer ramus about two-thirds as long as the peduncle and nearly as wide;
second joint slightly longer than the peduncle, equal to the first in width; and the

254 'BULLETIN OF THE UNITED STATES FISH COMMISSION.

terminal claw sinuate or irregularly curved. The stout seta on the outer margin of
the second segment of this ramus is borne at about a quarter the length of the seg-
ment from the distal end, and is approximately half as long as the segment to which
it is attached. The inner ramus is a little longer than the basal joint of the outer.
It is not dilated or otherwise modified, but terminates bluntly, bearing at the tip a
covering of long cilia.

The right antenna of the male is without notable distinctive characters. The
antepenultimate segment is as long as the two following taken together; the fourth
from the tip bears two long sword-like spines at its margin, both attached to its basal
fourth ; the expanded segments are well armed with conical spines, straight and curved,
but without hooks.

Small lakelet near Gardiner, Montana.

Epischura nevadensis, var. columbiae, u. var. (Plato XLI, Figs. 19-21.)

It is with pleasure that I report here the occurrence of another form of this inter -
esting genus of North American entomostraca, the fourth or fifth thus far discovered.
The first species described, E. lacustris, has been found in the Great Lakes, in the smaller
lakes of Wisconsin and Minnesota, and at Portland, Oregon; the second, E.fluviatilis
Herrick, has been seen only by the original describer of the species, by whom it is
said to occur in Mulberry Creek, (Jushinan County, Alabama.* Epischura norden-
sTtioeldii Lillj., is from Newfoundland, and E. nevadensis from lakes Echo and Tahoe,
the former in California, the latter partly in that State and partly in Nevada. The
present form occurs in Swan and Flathead lakes, in northwestern Montana, where it
was the most abundant copepod in the open water.

The absence of all representatives of this genus from the lakes of Yellowstone
Park, evidently adapted to them, hints strongly at a limit of altitude to their distribu-
tion. The highest locality from which any species has been reported is Lake Tahoe,
said to be 0,250 feet above the sea; while the lowest lake of suitable size in Yellow-
stone Park from which our collections were made was 1,200 feet higher than this.
This topographical difference does not measure the biological difference, however, as
the lower location is also more than five degrees south of the Yellowstone lakes.

Disregarding the doubtful fluviatilis, the species of Epischura are, so far as known,
of north temperate range in North America. The form least modified, both in abdomen
and fifth legs, is the Newfoundland species, nordenskiceldii, and the most modified in
both is nevadensis, lacustris standing intermediate. The new form, again, is interme-
diate between lucmtris and nevadcmix proper and may be roughly characterized as
uniting the characters of the fifth legs of the male and female and the caudal set* of
nevadensis with those of the abdomen of the male of lacustris.

* Herrick's species hardly seems to belong to this geuus. Tho abdominal processes are described
as projecting from the left side of the abdomen, and consequently can not be homologized with those
of E. lacustris. all of which are developed from the right, and the fifth foot of the male is scarcely
capable of close comparison with the corresponding appendages of undoubted Epischura. The differ-
ence reported in the position of the hinge in the antenna) of the male also points to a deep-seated and
fundamental distinction, not easy to reconcile, it must be admitted, with the agreement in fluriatilis
and laciistrii with respect to the inner ramus of the swimming legs, the fifth legs of the female, and
the caudal setae.

AQUATIC INVERTEBRATE FAUNA OP WYOMING AND MONTANA. 255

It differs frofli typical nevadensis in the more complete segmentation of the
ceph.ilothorax in both male and female, four sutures extending distinctly across the
back. It is also a little larger, adult females measuring 2.12 to 2.4 millimeters in
length to the tips of the rami, with an ordinary width across the cephalothorax of G.4
millimeters. The male is somewhat smaller, about 2.1 millimeters in length.

The antennaj of the female are long and slender, reaching about to the posterior
end of the penultimate segment of the abdomen. The first segment of the female
abdomen is as long as the two following together, and the furca is as long as the pre-
ceding segment. The female abdomen is not curved as in lacustris, and the sperma-
tophore extends downward and backward instead of curving upward, as in that species.
The three caudal setaj are all similar and of equal width, the base of each being a third
the width of the end of the furca. There is a short, stout, conical spine at the outer
distal angle of each ramus, and a small, soft seta at the inner angle.

The fifth legs of the female are broader in proportion to their length than in
lacustris, but more slender than in the nevadensis of Lilljeborg. The last segment
of the leg is four times as long as broad and bears six teeth (occasionally seven), four
of which are terminal. The inner of these four is commonly the largest, although the
third from within may equal it. The inner lateral tooth is close to the inner terminal,
and nearly or quite equals it in size, and the outer lateral is nearly opposite. Some-
times there are two teeth on the outer margin of this segment. The middle joint of
this leg is less than half as wide as long, and the basal is longer than wide.

In the male abdomen there are five distinguishable segments, as in all the other
species,* the second, third, and fifth bearing lateral processes extending to the right.
The first three segments are subequal in length. The lateral process of the second
has the form of a stout but thin wing or lamina projecting laterally a distance equal
to the width of the segment. It springs from a broad and thick prominence of the
segment itself; is acute at the apex, with the point a little recurved, convex and
smooth in front, as seen from above, and nearly straight behind, except that this edge
is irregularly and minutely serrate throughout and deeply emarginate where it joins
the segment. As seen from the side this blade is strongly curved downward (ven-
trally), like the following. The third segment bears a broad, thin lamina which projects
outward and a little backward from its posterior angle as a flat process, as wide as
long, curved downward and broadly rounded at the end, quite simple, except that it
is strengthened beneath by a ridge of chitin. These processes are in strong contrast
to the corresponding ones of nevadensis proper. From the fourth segment spring
two processes, the ventral of which is very similar to that of nevadensis, but broader, a
triangle in form, with nearly equal sides, with the apex slightly truncate and bearing
three serrations, and with the posterior side very minutely roughened. The dorsal
process of this segment is a small irregular plate curving forward, inward, and
downward.

Fifth pair of legs substantially as in nevadensis Lillj.

Abundant in Swan and Flathead lakes, Montana.

*A recent study of the male abdomen in lacuxtris shows beyond a doubt that the fourth and
fifth segments are flexibly articulated and that the fourth is without process, the fifth bearing two
processes, as in all the other species.

256 BULLETIN OF THE UNITED STATES FISH COMMISSION.

ROTIFERA.

Monostyla ovata, n. sp.

Lorica broad ovate, truncate, antennae nearly as broad before as behind. Dorsal
outline regularly rounded, not recurved before. Lobe flat; ventral plate flat. Toe
with very distinct shoulder, however viewed. Front margins of both dorsal and
ventral plates entire. Dorsal plate strengthened by two diverging longitudinal ribs,
about equidistant from each other and from the lateral angles of the plate, rendering
the anterior margin slightly angulate where they join it and vanishing behind at
about the middle of the shell. Eye single, transverse, oval, red, situated just above
and before the mastax, with two very minute red points behind it. Foot and toe
about two- thirds as long as the lorica is wide. Ventral plate much shorter posteriorly
than dorsal, its posterior margin slightly excavated before the foot.

Dimensions, 0.25 millimeter long by 0.18 millimeter wide.

From warm spring (103° P.), Yellowstone Lake, August 3, 1890.

Conochilus leptopus, n. sp.

Resembling G. volcox. Antenna? adnate to the tip, where the pair are rounded
off as one; but very slightly bifid at the base of the two hairs. Stalk not swollen,
slender, tapering backward regularly from the slightly dilated point of attachment to
the body. Eyes black, about midway between cloaca! opening and edge of disk.
Cloaca about half of distance from the edge of the disk to the base of the expanded
body. Trophi slightly tiuted yellowish brown, not orange.

Entire length, when expanded, 0.32 millimeter; extended stem, 0.13 millimeter;
breadth of body, 0.088 millimeter; expanded disk, 0.094 millimeter.

Exceedingly abundant in spherical colonies in Yellowstone Lake, July and August,
1890, and also in Lewis Lake, Yellowstone Park, July, 1890.

PROTOZOA.

Stentor igneus, var. fuliginosus.

Form turbinate, with slightly swollen sides, very slightly changeable, peristome
with spiral lobe, greatest width slightly more than half greatest length. Color soot-
black, given by blackish granules; when decolorized showing green as if by chloro-
phyl.

Individuals when highly magnified gray by reflected or transmitted light, through
mingling of green and black.

Form not symmetrical, the right side (when spiral lobe of peristome is uppermost)
being swollen below. When the peristomal lobe is at the side it appears as a tubercle
or projection. Form sometimes considerably shortened, so as not to be longer than
broad. Peristome angularly produced, so as to form with tip of spiral lobe an equi
ateral triangle, giving the entire animal an angular or trigonal shape. Tip flexible
and contractile, transparent when extended.

Swarmed in Lake of the Woods August 20, 1890, forming soot-like collections as
a film on the surface among pond lilies (Nelumbium) and a discontinuous coating on
under sides of the same.

258 BULLETIN OF THE UNITED STATES PISH COMMISSION.

. EXPLANATION OF PLATES.

PLATE XXXVII.

*

Fig. 1. Daplinia pulex, var. pulicaria. Male.
Fig. 2. Daphnia dentifera.
Figs. 3, 4. Cypris barbatus.

PLATE XXXVIII.

Figs. 5-8. Cypris barbatus. (5) First antenna. (6) Second antenna. (7) Postabdomen. (8) Man-
dible.

PLATE XXXIX.
Figs. 9-12. Cyclops thomasi. (9) First leg. (10) Second leg. (11) Third leg. (12) Fourth leg.

PLATE XL.

Fig. 13. Cyclops thomaii, fifth leg.

Figs. 14-17. Cyclops capilliferus. (14) Adult female. (15) First leg. (16) Fourth leg. (17) Fifth
leg.

PLATE XLI.

Fig. 18. Cyclops capilliferus, antenna of female.

Figs. 19-21. Epischura nevadensis, var. Columbia. (19) Abdomen of male. (20) Fifth pair of legs

of male. (21) Fifth pair of legs of female.
Fig. 22. Diaptomus piscinas, fifth pair of legs of male.

PLATE XLII.

Figs. 23-25. Diaptomus skoshone. (23) Fifth pair of legs of male. (24) Fifth leg of female.
(25) Right antenna of male.

Figs. 26-28. Diaptomus lintoni. (26) Fifth leg of female. (27) Fifth pair of legs of male. (28) Ter-
minal segments of antenna.

Bull. U. S. F. C. 1891. Aquatic Invertebrate Fauna of Wyoming and Montana. (To face page 258 )

PLATE XXXVII.

Bull. U. S. F. C. 1891. Aquatic Invertebrate Fauna of Wyoming and Montana. (To face page 258.)

PLATE XXXVIII.

Fig. K.

_BulL U' S- F. C. 1891. Aquatic Invertebrate Fauna of Wyoming and Montana. (To face page 258.)

I

PLATE XXXIX.

Fig. 9.

Fig. 10.

Fig. 19.

Bull. U. S. F. C. 1891. Aquatic Invertebrate Fauna of Wyoming and Montana. (To face page 258.)

PLATE XL.

Fig. is.

Fig. 1C.

Fig. 13.

Fig. 17

Fig. 14.

Bull. U. S. F. C. 1891. Aquatic Invertebrate Fauna of Wyoming and Montana. (To face page 258.)

PLATE XLI.

Fig. 20.

Bull. U. S. F. C. 1891. Aquatic Invertebrate Fauna of Wyoming and Montana. (To face page 258.)

PLATE XLII.

Fig. 23.

Fig. 24.

Fig. 26.

Fig. 27.

Fig. 28.

Fig. 25.

IND EX.

Page.

Aerial sound, note on 215

Bridge Creek 227

Bridge Bay : 228

Discussion of the collections 213-214

Descriptions of species and varieties 240-256

Duck Lake 230

Explanation of plates 257

FlatheadLake 238-288

Flathead River system 236-239

Genera and species referred to :

Acroperus leucocephalus 222, 230, 232

Agabus 233, 234

Agrion 218,220,232,233,239

Agrionina? 219, 230

Alona 225,230,232,239

Allorchestes 216, 219, 221, 227, 235

dentata 214, 218

219, 220, 225, 230, 232. 233, 238, 239

inerJnis 225, 227, 228, 229

Amnicola 220,229

A mphipoda 218

Annelida 233

Aulostoma lacustris 2i8

Bosmina 216,227,238

longirostris 222

Casnis 222

Candona 230

Ceriodaphnia 214, 227, 232, 233

reticulata 232. 233

Chwtogaster 233

Chara 226,231

Chironoraus .214, 217, 218, 219, 220, 222, 225, 226, 227, 228, 231,
232, 233, 234, 235, 238, 239

Chydorus 216,218,231,232

gphsericus 225, 227, 229, 232, 233

Cladocera 225, 233, 240, 244

Cladophora. ; 215, 222

Clepsine 220, 225, 226, 227, 231, 234, 235, 239

elegans 220

ornata 232

Crelambus .• 219

Coleoptera 227,239

Colymbetes , 233

Conochilus 219, 220, 225, 227

leptopns 225,229,256

volvox 219

Oopepoda 233,247,255

Corethra 214,220,228,230,233,234

Corisa . . -214, 217, 218, 219, 220, 226, 227, 228, 232, 233, 235, 239

Crustacea 238

Culex 218

Cnlicida) 227

Page.
Genera and species referred to— Continued.

Cyclops 214, 216, 220, 225, 227, 231, 232, 233, 235. 238

capilliferus 248-249

gyrinus 216, 218, 222, 225, 230, 238, 239

minnilus 216,225,230,247

serratus 247, 248

serrnlatus 216, 218, 225, 229

thomasi 238,249-259

Cypris 216, 218, 225, 226, 227, 228, 232, 239

barbatus ; 227,228,244-246

Rrandis 246

Daphnella 220, 227

brachyura 221 222

Daphnia 214, 218, 227, 231

angulifera 218,220

arciiata 222

clatbrata 235,240-241

dentata „ 244

dentifera 240, 243-244

pule* . . .214, 216. 218, 225, 226, 227, 228, 230, 232. 233,
237, 238, 242, 243

pnlez var. pulicaria 242-243

schosdleri 231, 232, 233, 234

thorata 237. 238, 239, 241, 242

1 >a pli Jii iil;i • 235

Dcronectes 226, 228, 233

griseostriatus 234

Diptera 218

Diaptomus. . .216, 217, 218, 219, 220, 221, 222, 225, 227, 228, 229

233, 235, 238

leptopns 253

lintoni 216. 225, 230, 231, 232, 233, 234, 235,

252, 253

piscinas 232, 253, 254

shoshone .214, 216, 219. 221, 225, 229. 232, 233, 235,

251-252

sicilig 216, 219, 222, 225, 226, 229, 232

stagnalis 251,253

Difflugia 222

globulosa 230

Dytiscidffi.. 218,232,233,238,239

Echinopyxis 222

Ephemera 220

Ephemerida) .227,239

Epischura 220

tliiviatilis 254

lacnstris 254, 255

nevadcnsia Columbian 238, 239, 254-255

nordenskiceldii 254

Eurycercus 232,238

lamellatus 218,222,227,239

II

INDEX.

Page.
Genera and species referred to— Continued.

Gammarus 216, 217, 218, 221; 225, 226, 227, 230, 232, 233,

235, 238

robustus 216, 218, 219, 228, 229, 239

Graphoderes fasciaticollis 230

Gregarinas 216

Haliplns 228, 230, 235

Holopedium gibberum , 219,222,234

Hydaticus 232

Hydrachnidao ..218,220,228,239

Hydra fusca 220,222

Hydrophilida) 238,239

Hydrophilus '. 217, 220

Hydroporus 232,233

Hygrotrechus 228.231

Lacinularia socialis 214

Leptodora 238

hyalina 222,239

Libellula '. 232

Libellulida 217

Libellulina? 230

Limnsa 227, 228, 229, 232, 233, 238, 239

Macrothrix 227,232

Melaniidas 220

Monostyla '.. 228

cornuta 229

ovata 256

Naidomorpba 230

Nelnmbium 256

Nephelia 226,235,239

4 strifita 219, 220

maculate. . . .218, 220, 221, 226, 228, 230, 232, 234, 239

obscura '- - . 216

Neuronia 217,221

Notonerta 217,218,219,231,232,233

Nuphar 218,220

Oligochasta 219,226

Oscillaria 229

Ostracoda 244-246

Paludinida? 220

Physa 217, 218, 219, 220, -J22, 226, 227, 228, 229, 233, 235,

238, 239

Phryganeida) 216, 233

Pisidium . . . .217, 218, 219, 220, 222, 226, 229, 231, 232, 234, 235,

238, 239

Planorbin 220,229,232,233,238,239

exacutuB 227

Plumatella 238,239

Podurida) 228,230

Polyphemus , 218,227,228

pedicnlns 216, 220, 222, 225, 228, 229, 232

Polyzoa 220

Potamogeton 215, 222

Pristina lacustris 238

Protozoa 256

Page.
Genera and species referred to— Continued.

Rotifera 256

Salmo mykiss 228

Scapholeberis . . ., 238

mucronatus 218, 225, 229, 230, 232

Sida 238

crystallina 220, 230, 231, 238, 239

Simocephalus 220

vetulus 227, 230, 231, 233

Simulium 228,229

Spha'riuni 232

SpongiUa '. 217, 226, 232, 234

fragilis 239

Stentor , 230

igueua 230

var. fuliginosus 256

Stylarialacustris 222

Tabanida? 227

Turbellaria 233

Unionidse 220

Valvata 220,228

Vaucheria 222

.Gardiner Lakelet 232

Gardiner River System 231-233

Ore be Lake 234-235

Heart Lake 220-222

Introductory 207

Lake of the Woods 230

Lewis Lake 218-220

Madison River System 234-235

Mary Lake 234

Persons referred to :

Birge.E. A 244

Boutelle, F . A .* 210,211

Brode, H. S « 212

Evermann, B. W 207,212,213

Harwood, E. L 207

Hofer, Elwood 208,211

Jordan, David S 207

Linton, EdVin 207, 208, 209, 218

Lucas, E. R 210,231

McDonald, Marshall 207

Parchen.W.H 212

Pelican Creek 227

Shonhone Lake 214-218

Small ponds, collections from. 232-233

Snake River System 214-222

Soda Butte Creek 229

Swan Lake 231-232,239

Trip of 1890 208-211

Trip of 189) 212-213

Twin Lakes 231

Yellowstone River System 223-230

Yellowstone Lake 223-229