SH 348
E85
Copy 1 ep eis
ed a ie
U. S. COMMISSION OF FISH AND FISHERIES,
JOHN J. BRICE, Commissioner.

A REPORT

4 Fans! 5 —~
ope hy P4228).

is 2
or
3 aN
= 4
i

UPON Gi JAN 12 1838 if
Ne! oF
LEC EIVEDY

SALMON INVESTIGATIONS IN THE COLUMBIA RIVER BASIN AND
ELSEWHERE ON THE PACIFIC COAST IN 1896.

BARTON WARREN EVERMANN and SETH EUGENE MEEK.

Extracted from U, 8, Fish Commisston Bulletin for 1897, Article 2, Pages 16 to 84, Plates 1 and 2, and 6 text figures,
Date of publication, January 6, 1898,

WASHINGTON:
GOVERNMENT PRINTING OFFIOE,.
1898.

U. S. COMMISSION OF FISH AND FISHERIES,
JOHN J. BRICE, Commissioner,

ok PORE

UPON

SALMON INVESTIGATIONS IN THE COLUMBIA RIVER BASIN AND
ELSEWHERE: ON THE PACIFIC COAST IN 1896.

BY

BARTON WARREN EVERMANN and SETH EUGENE MEEK.

Extracted from U, 8, Fish Qommission Bulletin for 1897. Article 2, Pages 15 to 84, Plates 1 and 2, and 6 text figures,
Date of publication, January 6, 1898,

WASHINGTON:
GOVERNMENT PRINTING OFFIOE.
1898.

2.—A REPORT UPON SALMON INVESTIGATIONS IN THE COLUMBIA RIVER
BASIN AND ELSEWHERE ON THE PACIFIC COAST IN 1896.

By BARTON WARREN EVERMANN anp SETH EUGENE MEEK.

During the season of 1896 the United States Fish Commission conducted a number
of investigations in Idaho, Washington, and Oregon, having for their general purpose
the improvement and extension of the fish-cultural operations of the Commission on the
Pacifie Coast. The specific object of each inquiry, the details of the work, and results
accomplished are given in the following pages. The work was under the imme-
diate direction of Professor Evermann, assisted by Dr. S. E. Meek, associate professor
of biology and geology in Arkansas State University; Mr. Ulysses O. Cox, professor of
biology in the State Normal School at Mankato, Minn.; Mr. A. B. Alexander, fishery
expert of the Fish Commission steamer Albatross ; Mr. W. F. Hubbard, superintendent
of the United States fish-hatchery at Clackamas, Oregon, and Mr. Alfred G. Maddren.

Investigations were carried on in ten different regions, as follows: (1) At the
Redfish lakes in Idaho, by Messrs. Evermann, Meek, and Maddren; (2) at Lake Pend
@Oreille, Idaho, by Messrs. Alexander and Cox; (3) at Wallowa Lake, Oregon, by
Messrs. Meek and Maddren; (4) on tributaries of the Lower Columbia, by Messrs.
Evermann, Meek, and Hubbard, and by Messrs. Alexander and Cox; (5) on streams
tributary'to Puget Sound, by Messrs. Alexander and Cox; (6) at Lakes Washington,
Sammamish, and Union, by Mr. Alexander; (7) on the Siuslaw River, Oregon, by Dr.
Meek; (8) on Whoahink, Tsiltcoos, and Tahkenitch lakes, Oregon, by Dr. Meek; (9)
at Upper Klamath Lake, by Messrs. Meek and Alexander; and (10) at Crater Lake,
Oregon, by Messrs. Evermann and Cox.

During the progress of this work the representatives of the Commission were the
recipients of material assistance and many acts of courtesy from various citizens, to all
of whom we are glad to acknowledge our indebtedness and to express our thanks.
Especial mention should be made of Mr. G. H. Stevenson, Vancouver, Wash.; Seufert
Brothers and Mr. I. H. Taffe, salmon-canners at Celilo, Oregon; Mr. Frank Warren,
salmon-canner at Portland; Captain Steers, of the steamer Lillian, Florence, Oregon ;
Mr. William Kyle and Messrs. Elmore & Sanborn, salmon-canners at Florence; Mr.
Leonard Christianson, of Acme, Oregon, and Mr. Scott Morris, of Ada, Oregon.

THE REDFISH LAKES, IDAHO.

It was found in 1894 that these lakes and their connecting streams afforded unusu-
ally good facilities for the study of the spawning habits of the chinook salmon and
both forms of the so-called redfish. The matter was taken up in the following year
and a continuous series of observations was carried on at those lakes from July 17 to
September 24. A detailed account of those observations, together with full descrip-
tions of the physical features of the region, may be found in the published report, in

the Bulletin of the United States Fish Commission for 1896, pp. 149-202.
15

16 BULLETIN OF THE UNITED STATES FISH COMMISSION.

The investigations of 1895 resulted in the following conclusions coneerning a
number of important questions regarding these fishes, viz:

1, The chinook salmon resorting to the headwaters of Salmon River for spawning
purposes reach the spawning-grounds in perfect condition, so far as shown by external
appearances, no mutilations nor injuries of any kind haying been seen. The many
sores and mutilations seen upon them later in the season result from injuries incident
to spawning, are received on the spawning-beds, and are not due to injuries received
en route from the sea.

2. After spawning, the chinook salmon coming to these waters die in the vicinity
of the spawning-beds.

3. The young chinook salmon remain about one year in the streams where they
were hatched before beginning their journey to the sea.

4. The large redfish are identical with the blueback salmon, or sockeye (Oncorhyn-
chus nerka), and come up from the sea, as do the chinook salmon. They reach the
spawning-beds in perfect condition, are mutilated there during the spawning season,
and then die, never returning to the sea.

5, The small redfish spawn at the same time and on the same beds with the large
redfish. They arrive upon the spawning-beds in apparently perfect condition, but soon
become mutilated, just as do the large form and the chinook salmon, and then die
without returning even to the lake.

6. Eggs of the small redfish laid in September in the inlet to Alturas Lake began
hatching on the 21st of March following.

7. The young redfish remain in Alturas Lake at least one year from the time when
they were spawned.

The investigations of 1895 left unanswered, however, some of the most important
and interesting questions concerning the redfish. It was not possible to determine
when the large redfish arrives at the Idaho lakes, nor whether the small form is really
anadromous. Both forms were already in Alturas Lake before the nets were set in
the outlet, on July 20. The importance of settling these questions, if possible, and
the desirability of repeating and verifying the observations of 1895, justified the
continuance of the work another season. .Alturas Lake was, therefore, again visited
in the summer of 1896. Camp was established July 11 on the outlet of Alturas Lake,
at the ford about 14 miles below Perkins Lake. On account of unusually high water
it was impossible to set the gill nets satisfactorily until July 17. On July 22 the camp
was moved to the head of Alturas Lake and a gill net was placed across the inlet at
its mouth. The net at the ford was taken up August 6 and reset in the outlet between
Perkins and Alturas lakes, thus making it more easily reached from the camp. It
remained in this place until September 25, when it was taken up and not reset again.
Redfish began to come into the inlet August 3. After that date the net was kept in
place only at irregular intervals until August 14, when it was taken up.

High water.—The spring of 1896 was an unusually late one in Idaho. The snows
of the preceding winter were much deeper than usual and were slow in disappearing in
the spring. Rains were also uncommonly frequent, even late in the spring and during
the first half of summer, the last of importance occurring August 6. A very hard
rain fell on the night of August 3, resulting in a perceptible rise in the streams. On
July 11 the streams were much higher than at the same time the year before. Salmon
River and Alturas Creek were overflowing their banks in many places; much of the

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. NY

valley above Alturas Lake was from a few inches to 3 feet under water. Not until
about July 22 was it possible to reach the head of the lake with a wagon, and the
outlet could not safely be forded earlier than that date. After July 14, however, the
water ran down very rapidly, and during August and September Alturas Creek was
only slightly higher than in 1895.

Temperatures.—Temperature observations were made as regularly as circum-
stances permitted. The records for July 11 to July 21, inclusive, are those made at the
camp on the outlet of Alturas Lake. The thermometer for air temperatures was
kept hanging in the shade on a tree at the camp; the water temperatures were taken
in the creek at the camp in water about a foot deep.

Temperatures recorded at the outlet of Alturas Lake, Idaho, July 11 to 21, 1896.

le Saar ]
Date. | Hour. Air. | Water. Date. Hour. Air. | Water.
| |
| oF, oF, oF,
July 11 62 58 July 17 55
12 | 59 56 | 60
| 7 58 | 59
52 55 | 18 58
13 | 52 53 | 63
| 60 55 60
| 55 53 19 58
14 | 58 53 | 62
70 58 61
62 54 20 58
15 63 54 63
71 57 61
| 61 56 21 58
16 60 60 63
74 59 6L
64 56

The temperatures recorded in the following table are those taken at the camp at
the head of Alturas Lake. The “inlet” temperatures were taken a few feet above the
lake where the water was about 2 feet deep; those for the lake were taken just in
front of our camp at a depth of about 3 feet, and those of the air were in the shade:

Temperatures recorded at the head of Alturas Lake, Idaho, July 22 to August 5, 1896.

Water. | | | Water.

Date. Hour. | 2S | ————= | Date. | Hour. 220) | = pe

| | Inlet. | Lake. | Inlet. | Lake.
0k aya hh AeA ea oF, | oF. | oF
July 22} 8a.m......... GY eae oec 565 || July 29 | 1 p.m.-..-.-.. 65 48 59
12m... os CA Pee 57 || Spates =e 54 50 56
8 p.m. 60 51.5 50 30) |.8ia. m.-.---..- 49 43 56
23 | 8a.m.- 55 | 48 53.5 DAs atts 68 49 60
12m ... 64 47 54.5 Liebe eee aae ce 55 52 58
9p.m.- 48 49 51 | 31 | 8a.m....--... 53 H 56
24| 8a.m. 46 41 53 Wp ss a= ae 74 50 59
12m... 64 47 55 | Tene = A255 59 51 55
8 p.m. 56 51 51 Aug.) Ui) Sia. me-. = -- 2. } 59 44 58
25 | 8a.m.. 52 42 54 i yeoroascetese 74 52 54
12m .:. 69 48 50 iyi po Se eae 59 53 55
8 p.m.- 53 | 51 51 Pe) | ISS es Soe aee 57 44 57
26 | 8a.m.. 46 | 42 50 LU Rae Ree 66 51 58
12m .- 63 47 59 | 8p.m......... ; 53 53.5 53
8 p.m 53 51 56 3 | 8a.m_....._.. 49 44 57
27 | 8a.m 50 43 54 12m 70 49 57
12m 64 47 58 8 p.m 42 52 50
8p. 60 50 59 4) 8a.m 41.5 42.5 55

28 | 8a. 48 55 54 12m... 57 50 51.5
12m 57 51 56 8 p.m 44 51 55
8p. 49 49 56 5 | 8a.m 42 42 55
29) 8 a.m 52 43 56 12m... 60 44 56

a During the night of August 3 there were heavy thunder and rain, snow falling on the mountains,
F. C. B. 1897—2

18 BULLETIN OF THE UNITED STATES FISH COMMISSION.

In the following diagram are shown the temperature curves obtained by taking
serial temperatures in Alturas, Big Redfish and Wallowa lakes. Two series were
taken in Alturas Lake and one in each of the others.

EIURARERR eos

ane

oak
el Ca : Te ae
a

Beets elale rele
=ilial

Hf
7
LY
a
20 vie
a)

N
°o

PELE TLL] © orererence tase.

——————“I

ie)
o!1o

ALTURAS LAKE,IDAHO.
AUG.I17%1896,BET. 3&6 PM.

co}

BIG REDFISH LAKE,IDAHO.
AUG. 971896, BET. 2:30 &3:30 PM.

oO;oO;oO

WALLOWA LAKE.OREGON.
AUG. 20™, 1896.

nN
10.0)
Ei

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896 is)

The following temperatures were taken August 11, in Alturas Lake just beyond
the bar at the mouth of the inlet. AIl are bottom temperatures except the first three.
The first, second, third, and fourth were all taken at the same place, the others a short
distance away:

ALG SL a Cae ee etree 57° 5: At depth) ofi72iteet----=----------- 42°
2. At depth of 24 feet -.-...--..---.-.. 54 6. At depth of 96 feet..........-.----- 404
3. At depth of 48 feet .....-..--.-:.--. 45 7. At depth of 102 feet..---2.--.---.-. 404
4. At depth of 90 feet -....-..----.---- 403

Two series of temperatures at different depths were taken in Alturas Lake August
11, between 3 p.m.and 6 p.m. These are shown on the temperature diagram on pre-
ceding page. The series indicated by the crosses shows the- bottom temperatures,
obtained by starting near shore in 1 fathom and going outward into increasing depths.
The circles on the continuous line indicate a vertical series taken near the middle of
the lake, where the depth was 150 feet. It will be noticed that, after reaching a depth
of 30 feet, the two series coincide exactly and that the temperature at the bottom in
the deepest part of this lake is but slightly, if at all, above that of fresh water at its
greatest density, namely, 39°.

A similar series was taken at Big Redfish Lake August 9, between 2.30 and 3.30
p. m., the results of which are platted on the same diagram, an examination of which
shows that this lake is, as a whole, considerably colder than Alturas Lake. Though
the surface temperatures are nearly the same, the water in Big Redfish Lake is seen
to grow colder much more rapidly as the depth increases. This is apparently due to
the greater size and depth of Big Redfish Lake, the lower temperature of the inflowing
water, and the greater protection afforded by the higher surrounding mountains. In
both lakes, however, the temperature decreases rapidly until a depth of about 100 feet
is reached, and beyond that depth the decrease is less than 2° in either case.

These temperatures were taken with a Negretti-Zambra deep-sea thermometer
tripped by means of a messenger, and the results are believed to be reliable. The
bottom temperatures which we took at these lakes in 1895 were taken with a Wilder
protected thermonieter, and we are now convinced that the results obtained are wholly
unreliable whenever the depth exceeded a few feet. The thermometer can not be
drawn up quickly enough to prevent the mercury from being warmed by the warmer
water above. The few bottom temperatures given in the report of the work done on
these Idaho lakes in 1895 are therefore worthless.

SUMMARY OF OBSERVATIONS REGARDING THE REDFISH.

Camp was established on the outlet of Alturas Lake July 11, and observations
began on that date. The daily inspection of the nets and the periodical examination
of Alturas Creek and Salmon River were carried on essentially as during the season
of 1895. Although gill nets could not be well placed until July 17, the opportunities
for examining the stream make it certain that if any redfish had ascended to the lake
between July 11 and 17 they would have been seen.

Beginning with July 17, gill nets were kept constantly in the outlet of Alturas
Lake, set in such manner as wholly to obstruct the stream to the ascent of fish. These
nets were kept in place until September 25, but not a single redfish, either of the
large or small form, was caught in them. No redfish were seen in the outlet or else-
where below Alturas Lake during the season. Small redfish appeared in the inlet in
considerable numbers in August and September, but no large redfish were seen this
year at any of the Redfish Lakes.

20 BULLETIN OF THE UNITED STATES FISH COMMISSION.

The first redfish seen was caught in Alturas Lake near the inlet July 16, while
fishing for Dolly Varden trout in about 30 feet of water. It was a male, 114 inches
long, and was taken on a hook baited with salmon spawn. Its stomach was well filled
with food, chiefly small crustaceans. Its color was a rich metallic blue on the back,
becoming silvery on the lower sides and under parts. The reproductive organs were
not greatly developed.

The next redfish were caught in the inlet gill net on the night of the 2d of August,
3 males in prime condition. On the next night 3 more males were taken, in excellent
condition and nearly or quite ripe. On the next night 8 males and 3 females were
caught, all the males ripe and the females nearly so. No others were seen until
August 6, when about 40 were observed in a deep hole about a mile above the lake.

As the redfish were now running up in considerable numbers, the nets were taken
up and not put down again except for one night, August 13, when 53 males and 32
females were caught. None of these showed any sores or mutilations of any kind.

By the middle of August the fish had arrived in large numbers, the maximum
being reached during the last week in that month. The fish in the inlet were counted
at intervals and a careful record made of their numbers. The record reads as follows:
1,044 fish on August 18; 1,345 on August 21; 1,038 on August 25; 1,558 on August
28; 1,354 on September 1; 1,515 on September 4; 1,286 on September 9; 1,067 on
September 12; 952 on September 14; 703 on September 18; 214 on September 25; 106
on September 28; October 2, none.

The fish seen on the first few days were all in perfect condition, but soon after
they began spawning mutilations appeared, and some died as early as August 23.
After that date they died rapidly, but the numbers were reinforced from time to time
by new schools which continued to come up from the lake until probably the first week
in September. The large decrease shown August 25, compared with the number
observed four days earlier, was apparently caused by a great many fish being caught
out by campers. Beginning early-in September the number of dead fish increased
and the number of live ones gradually decreased until October 2, when all had died.
The dead fish were usually found in the deeper pools or quiet portions of the inlet.
On September 14 one dead redfish was found lodged against the net in the outlet a
few rods below Alturas Lake, and on September 25 four were found washed up on the
shore of the lake a few rods west of the inlet. These had all doubtless been carried
down by the current.

Run later than in 1895.—In 1895 the first redfish appeared in Alturas Inlet July
25, the maximum was reached about September 6, and all had died by September 25.
In 1896 the run began nine days later, the maximum was attained about nine days
earlier, and the last live fish were seen about nine days later than in 1895. Though
the season began a little later, it covered almost exactly the same length of time.

Mutilations.—Previous observations concerning the cause of the mutilations were
verified. As the fish came into the inlet they were all in perfect condition, and con-
tinued and careful observation of their movements during the spawning season showed
conclusively that the injuries are caused by contact with the gravel and by fighting
on the spawning-beds.

Dying of the redfish.—The conclusion reached last year, that every redfish dies
after spawning once, was based upon absolute proof, so far as this region is concerned,
and no other conclusion was possible from the observations made this year. The

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. all

number of redfish entering Alturas Inlet in 1895 was about 2,000 in round numbers,
and not one ever returned alive to the lake below. The run this year was somewhat
larger. The greatest number counted at any one time was 1,558, but many had died
or been caught by campers prior to that time, and many new schools came in subse-
quently. It is a conservative estimate to say that between 2,500 and 5,000 fish entered
Alturas Inlet in 1896. While the observations for the purpose of determining whether
any returned to the lake were not as carefully made as in 1895, there is no evidence
that a single fish ever returned to the lake alive.

While the mutilations are usually quite severe, in many cases they are not par-
ticularly so, and unmutilated fish were sometimes found dead or dying. This fact
seems to be of great importance because of the light which it throws upon the probable
cause of the death of spawning salmon. Some naturalists have maintained that the
dying is attributable to the injuries; others that exhaustion resulting from the long
journey from the sea or long abstinence from the use of food is the cause; but as a
matter of fact, many observers have seen salmon dying in large numbers at the end
of the spawning season in waters only a few miles from the sea, and even in salt
water, and the cause could, therefore, not have been exhaustion resulting from long
journeys without food, or mutilations received on such journeys. That the dying
results from long abstinence from food is completely disproved by observations at
Alturas Lake, where redfish have been observed to come up out of the lake with food
in their stomachs, and have died a few days after spawning. The true cause is evi-
dently deeper seated in its nature and more general in its application than has been
supposed. :

No large redfish seen in 1896.—In 1894, 14 large redfish were seen in Alturas Inlet
and 1 in the inlet to Pettit Lake, and as these waters were visited only once this does
not, of course, indicate the total number that may have spawned there in that year.
In 1895 only about 8 large redfish were seen, 3 in the inlet to Pettit Lake, 3 in Alturas
Inlet, and 2 in Alturas Lake. During the season of 1896 no large redfish appeared in
Alturas Inlet, nor were any seen at Pettit Lake, which, however, was not visited
sufficiently often to fully determine the matter. None was observed at Big Redfish
Lake during a trip there August 8 to 10, and parties who visited the lake during
August and September report seeing no redfish.

Are the small redfish anadromous?—This question can not yet be positively
answered. If they are, they had reached Alturas Lake prior to July 11, just as they
must have reached it before July 20, in 1895. But this may very well be so, for the
big redfish, which is undoubtedly anadromous, had also reached this lake earlier in 1895
than July 20. Observations at Alturas and Wallowa lakes point strongly to the
probability of the small redfish being permanently resident in those lakes. The first
of these was the catching of a small redfish in Alturas Lake July 16. This fish took
the baited hook, and its stomach was found to contain some food, chiefly insect larvie
and small crustaceans. One caught in the gill net August 6 had a trace of food in its
stomach. Twelve small redfish were caught with grabhooks in Wallowa Lake by Mr.
J. J. Stanley, about the first of September, and the stomachs of 9 of them were
found to contain food consisting almost wholly of entomostracans and other small
crustaceans. In one case the stomach contained a small quantity of some alga and
in five or six cases the stomach was quite full of food.

22 BULLETIN OF THE UNITED STATES FISH COMMISSION.

The twelve specimens taken by Mr. Stanley were males which probably would
have spawned that season. They are considerably smaller than any heretofore taken
and vary from 62 to 92 inches in total length; one found dead on the shore of Wallowa
Lake August 24 is only 54 inches long.

In the light of all these facts it is hard to believe that the small redfish come up
from the sea. On the other hand, when we consider that the large and small forms
present no structural differences of value, that they spawn at the same time and on
the same beds, that the small form has never been seen, so far as known, in any of
these lakes except at spawning time, and that they, like other members of the genus,
die after once spawning, it is equally difficult to believe that they are permanently
resident in fresh water.

Chinook salmon in the upper Salmon River.—The number of chinook salmon coming
to these spawning-beds in 1895 was not far from 1,000. The first appeared July 24, and
the maximum nuniber was reached about August 24, All arrived in perfect condition,
and all died soon after spawning, none returning to the sea.

Not much time was given to the chinook salmon in 1896, only occasional visits
being made to Salmon River. From such observations as were made and from infor-
mation obtained from various parties who came up the river from Stanley Basin and
other points below, it appears that the run was very small and much later than in the
preceding year. Four were seen in Alturas Creek, at the mouth of Pettit outlet,
August 24, and seven days later 7 more were seen in the Salmon River a mile below
the mouth of Alturas Creek. On September 3, 8 were seen from this point down to
Roaring Creek. The same fish, or possibly others, were seen at various times until
September 11, after which no more were seen. Mr. Springer, who was fishing and
hunting for the Custer market, says that in Bear Valley and Sulphur creeks, where
salmon have until this year been plentiful, not one was seen. He saw only 1 salmon
in Salmon River below Roaring Creek.

LAKE PEND D’OREILLE, IDAHO.

In February, 1889, the United States ish Commission placed in Lake Pend
d@ Oreille 1,300,000 fry of the common whitefish (Coregonus clupeiformis). Until now
no investigation had ever been made to determine whether any of these fish survived.
Fugitive reports have come to the Commission from time to time of the capture of
examples of this species, but specimens submitted for identification proved to be the
common native species (Ooregonus williamsoni). If the plauted species succeeded in
establishing itself, the individuals would by this time have attained a size sufficient
to enable them to be taken in gill nets.

Mr. A. B, Alexander and Prof. U. O. Cox were instructed to go to Lake Pend
d’Oreille and endeavor to determine the result of the whitefish plant made in this
lake in 1889, and to make investigations regarding the suitability of the lake, in its
physical and biologic features, to the common whitefish. The investigations were
begun June 25 and it was hoped they might continue through the season, but, owing
to urgent work on Puget Sound, which Messrs. Alexander and Cox were directed to
perform, the Pend d’Oreille investigations were discontinued July 16. Excessively
high water prevailed during the time spent upon the lake and the conditions were not
favorable for investigations of this kind. Gill nets were set in different places in the
north end of the lake, but no tests were made in the southern portion. While no

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. Do

whitefish were found, the result can not be regarded as conclusive. The investigations
should be resumed and continued until the whole lake can be carefully examined.

The report upon the physical examination of the lake and upon the fish-food
collected will be reserved until a more thorough study of this important body of water
can be made.

WALLOWA LAKE, OREGON.

It has been known for several years that both the large and small redfish have
spawning-beds in the inlets of this lake, and that chinook salmon spawn in Wallowa
River and other streams in that region. A short visit was made to Wallowa Lake by
Messrs. Meek and Maddren on August 19, and they remained there and on Wallowa
River until August 26.

Wallowa Lake is in the northeastern portion of Oregon and inthe southern part
of Wallowa County. It lies in the southern end of a considerable valley, which is
bounded on the east, south, and west by the Powder River Mountains. This lake is of
glacial origin, and is bordered on the east and west sides by immense lateral moraines,
which rise about 1,200 feet above the surface of the lake at the upper end, but become
less high as the lower end of the lake is approached. Across the broad canyon at the
foot of the lake is a terminal moraine a few feet high, through which the outlet has cut
its way. At the foot of each of the lateral moraines, about 20 feet above the surface
of the water, is a bench from 20 to over 100 feet in width. This bench was evidently
formed when the surface of the lake was at that height. The outlet has cut a sort
of canyon through the terminal moraine and lowered the lake to its present level.
The sides of the moraines above and below these benches are very steep, aud except
along the upper one-third of the lake there is very little timber on them. There is a
dense growth of willows and alder along the south shore of the lake, extending back-
ward into the valley about one-fourth of a mile. The rest of the valley near the head
of the lake is quite heavily timbered.

The mountains near the head of the lake are very rugged, and are said to be over
5,000 feet above the surface of the lake.

Wallowa Lake is about 44 miles long from north to south, with a maximum width
of about 1 mile. The greatest depth found was 250 feet. Its shores are quite regular,
there being no marked indentations anywhere. The beach at the upper end of the
lake is sandy; that on the other three shores is for the most part covered by bowl-
ders. The inlet of this lake has its origin in some lakes about 3 miles farther up the
narrow valley. The largest of these is said to be about half a mile in diameter.
About 25 miles above Wallowa Lake are Wallowa Falls. Just before reaching this
point the inlet flows through a canyon about 90 feet in depth, the height of the falls
being about 50 feet. The water falls into a basin about 50 feet in diameter, the outlet
of which is 15 to 20 feet wide. A great many bull trout are caught with hook and
line in this basin each year. Between the falls and the lake the inlet flows most of
the distance, with a swift current, over a bed of bowlders. Within about half a mile
of the lake the inlet breaks up into two or more channels, and finally discharges its
waters into the lake by two mouths, which are only a few rods apart. The mouth of
the inlet has been changed within the past few years. The stream is very swift and
contains very few gravel beds, and these are along its lower mile. Another small

24 BULLETIN OF THE UNITED STATES FISH COMMISSION.

inlet empties into the lake near its southeast corner; this inlet is only a small brook
in summer. The lakes and the inlet above the falls are said to contain no fish life.

Wallowa River, as it leaves the lake, is a very rapid stream and flows over a bed
of bowlders. The first half mile of its course is through a canyon about 30 feet deep.
From Joseph to Lostine the river is said to be very rapid.

The West Fork of Wallowa River rises in the mountains west of Wallowa Lake,
and, after flowing in a general northerly direction until within about a mile of the
main fork at Lostine, it bends to the westward and joins the main fork a short dis-
tance below that town. About 12 or 15 miles below where these two forks come
together Wallowa River is joined by Minam River, which forms a part of the western
boundary of Wallowa County. The river here flows nearly due north, and soon
unites with the Grande Ronde River, a tributary of Snake River.

About 4 miles from Lostine up the West Fork of Wallowa River, the river was
narrower and flowed with considerable velocity among the huge bowlders scattered
over its bed. About 3 miles above Lostine there is a somewhat level stretch in the
river of about half a mile, which includes some gravel bars. From Lostine to the
mouth of Minam River the Wallowa is a rapid stream, flowing most of the distance
over a bed of bowlders. Gravel or sand bars are not common. The water is very
clear, and at the time of our visit was not over 3 feet deep in this portion of the stream.
Minam River is smaller and is said to be very similar to the Wallowa. The Grande
Ronde River was seen at Lagrande and Elgin and at various places between those
towns. It appears to be a rather sluggish stream, the water being warm and muddy.

Soundings and Temperatures.—The following soundings and temperatures were
taken on Wallowa Lake about half a mile north of the south end of the lake. We
began on the west side and counted the strokes of the oars as we crossed, taking
soundings and temperature at intervals, as shown below. The first sounding was
made 20 oar strokes from the west shore.

Soundings and temperatures taken in Wallowa Lake, August 20, 1896.

f
Strokes Strokes
taken Temper- taken ‘Temper-
between | Depth. | ature at | between | Depth. | ature at
sound- | bottom. sound- bottom.
ings. ings. |
| »
No Feet. F No. | Feet. oF
20 =| 42 51 50 219 402
100 216 404 30 180 402
100 | 236 402 30 99 414
100 | 234 402 35 strokes to east shore.
100 228 403 | |

Temperature of water at surface 65°.

The greatest depth, 250 feet, was about a mile from the south end of the lake and
about a third the distance across from the west shore; the bottom temperature here
was 404°. About a quarter of a mile south of this point and near the middle of the
jake from east to west a depth of 246 feet was found, and the same bottom temperature.
A series of temperatures taken August 20 is platted on page 18. The surface of the
iake near the inlet at 3 p.m., August 24, was 62° when the air was 70°. The tempera-
ture of the east inlet at the same time was 52° and the other was 48°. The water in a
small spring creek was 44°.

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 25

Redfish at Wallowa Lake—No large redfish were seen at the time of our visit to
this lake, August 19-26. From interviews with persons familiar with the region it
appears that the large redfish usually reach the lake during the last half of July. They
come up the river with the first run of chinook salmon, the run into the lake lasting
about three weeks. Soon after entering the lake they are seen in large schools at its
upperend., They begin to spawn in September. They are not red when they enter
the lake, but become so a short time before spawning. They spawn in the inlets of
the lake and on the shores of its upper end. They were formerly caught for the market
during the month of August. According to Mr. J. J. Stanley, of Joseph, the run of
large redfish in Wallowa Lake in each year since 1882 was as follows:

1883. The run of redfish was very large.

1884. It is estimated that less than 100 redfish entered the lake.

1885. About 75 were caught with a seine, the run being very small.

1886. The run was very large, but not as large as in 1883.

1887, 1888, and 1889. Very few redfish were caught with spears in the river above the lake.
1890. The run was quite large; two parties salted many for the home market.

1891, 1892, and 1893. Very few redfish seen in the lake during these years.

1894. About 2,000 redfish were caught at the head of the lake and salted.

1895. The run was again very small.

1896. The run was very small; about one dozen were seen in the lake and one in the inlet.

In the spring many small fish, from about 6 to 8 inches in length, are stranded in
irrigating ditches, many others are killed by entering mill-races. By the inhabitants
of Joseph these are thought to be the young of the big redfish on their way to the sea.
The large redfish are not known from any streams in this region except in the lake
and its outlet and inlets. It is said never to run up the West Fork of the Wallowa
River. Only one small redfish was observed while at Wallowa Lake, a small example
5} inches long, which was found dead on the lake shore August 24.

According to Mr. Stanley, the small redfish are seldom or never seen there except
in September, and then only in schools at the head of the lake or in the inlets where
they spawn, at the same time and on the same beds with the large redfish. He says
the small ones are called “yanks,” and that not more than 1 in 15 is a female; 12
which he caught for us were all males. According to Mr. Stanley, the fish which he
calls “‘grayling” is usually seen only in June, when it is easily caught by trolling in
the lake. The small redfish found dead on the beach of Wallowa Lake, August 24,
was called a grayling by Mr. Stanley, from which it seems probable, that-the so-called
“orayling” are the more silvery small redfish.

Chinook salmon.—This salmon is known to enter both the West and Main Forks
of Wallowa River, the majority running into the West Fork. Those which keep in
the Main Fork enter Prairie Creek, where they have their spawning-beds. They have
been seen in the outlet close to the lake itself. The early run isin July. The largest
run 1s in September when they are known locally as “dog salmon.”

West Fork was examined for a distance of 4 miles, that portion of the river most
frequented by chinooks, but no salmon were seen. Four had been caught, however,
on the spawning-beds 3 miles above Lostine a few days before, and one was killed by
an Indian below Lostine about the same time.

Steelhead trout.—The steelhead reaches Lostine on the Wallowa River in March
and April. They spawn mostly in the West Fork of Wallowa River, very few being
seen in the Main Fork.

26 BULLETIN OF THE UNITED STATES FISH COMMISSION.

LOWER COLUMBIA RIVER.

It being the desire of the Commissioner to establish a station for the hatching of
salmon in the Lower Columbia River basin, we were directed to visit and examine
various streams, particularly in the vicinity of The Dalles and Caseades, and select a
site suitable for such purposes.

The conditions requisite for such fish-cultural operations as were contemplated
are essentially as follows: (1) An abundant supply of salmon easily obtainable when
ripe, or nearly so; (2) water of proper quality as to purity and temperature and in
sufficient quantity; (3) suitable land upon which to locate the hatchery building, and
So situated as to permit the water to be brought to the hatching-troughs by means of
gravity; (4) proximity of building materials and good railroad facilities.

Upon taking up this inquiry it became at once evident that no definite informa-
tion existed concerning the location of salmon-spawning beds anywhere in the Lower
Columbia River basin and information regarding the spawning time was equally
uncertain and unreliable. It was therefore necessary to visit as many streams as
possible and determine the facts by personal inspection and by interviews with people
living on or near them.

The telegraphic instructions dated August 1 directed that a site be definitely
selected, and as soon as possible, in order that the station might be operated during
the season of 1896. In order to be able to examine as many streams as possible in
the brief time at our disposal Messrs. Cox and Alexander were ordered: from
Puget Sound to examine the Lewis River and other streams below Vancouver, while
Messrs. Evermann and Hubbard made examinations at the Cascades, The Dalles,
Hood River, Big White Salmon, and Celilo. These inquiries were made during the
first ten days of August. Later (August 28 to September 6) Messrs. Evermann and
Meek examined the Des Chutes, John Days, Hood, Big White Salmon, and Little
White Salmon rivers, and on September 7 and 8 Messrs. Evermann and Hubbard
reexamined Little White Salmon River and examined Tanner and Eagle creeks near
Bonneville. Mr. Hubbard also visited Hamilton and Rock creeks across the river
from Cascades, and Mr. Alexander made an examination of Toutle River, Wash-
ington. Only the more important results of these various inquiries are given here,
detailed reports having been made at the time to the Commissioner.

LEWIS RIVER.

This river forms the boundary between Clarke and Cowlitz counties, Wash., and
flows into the Columbia between Vancouver and Kalama. Two main branches, known
as the North Fork and South Fork, unite only a few miles above the mouth of the river,
The North Fork has its headwaters on the eastern slopes of Mount St. Helens; the
South Fork rises in Skamania County, southeast of Mount St. Helens. No salmon
were seen in the North Fork at the time of Mr. Alexander’s visit, during the first
week in August. Itis said that the salmon of the spring run do not enter the river, but
that large numbers of chinook salmon of the fall run enter it; also a good many silver
Salmon and steelheads. The chinooks are said never to run in until after the 10th of
August, when the close season begins. When the close season ends (September 10)
fishing with gill nets and seines is carried on to some extent in this river. Some
logging is carried on in this stream during the spring, but it ceases long before the
salmon arrive.

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 2Alh

Considerable spawning-beds are said to be located in the vicinity of tna. The
water is pure and cold, and if chinook salmon really come there in large numbers all
other conditions are favorable for fish-cultural purposes.

The South Fork was examined as far up as La Center. This stream is about the
same size as the North Fork, but the water is not clear. The banks, in the lower
portion at least, are composed of red clay or easily disintegrated rock, and the water
is thus usually kept more or less muddy. Below La Center there seems to be no
bottom suitable for spawning-beds and there are no tributary creeks from which water
suitable for hatchery purposes could be obtained. Chinook salmon of the fall run are
said to enter this fork, but the locality of their spawning-beds is not known. No
salmon were seen by Professor Cox at the time of his visit.

TOUTLE RIVER.

This stream is an eastern tributary of the Cowlitz, into which it flows a few miles
north of Castle Rock. About 11 miles above its mouth is a fall 15 feet high. From
this point down there are many riffles suitable for spawning-beds, and where racks
could be put in without difficulty. One of the best places is at the mouth of Pat’s
Creek, about 7 or 8 miles above the mouth of the river, where the stream is 35 to 75
feet wide and the bed is of coarse sand and gravel. Suitable water for hatchery
purposes could be obtained from the creek.

It is claimed that the fall run of salmon enters this river in considerable numbers,
including chinook, silver, dog, humpback, and steelhead. Salmon appeared to be
scarce at the time the river was examined, as only three were seen by Mr. Alexander;
but it was stated that they would be more abundant about the Ist of September. They
are said to collect in numbers at the foot of the falls already alluded to, and if this be
true, that would prove a good place for collecting the eggs. The falls do not seem
to offer an impassable barrier, as salmon are reported from points still farther up the
stream.

* CASCADE LOCKS.

On the Washington side of the Columbia River, opposite Cascade Locks and at
the head of what is known as the “Old Incline” at the Upper Cascades, is a place
where it was thought fish-eultural operations might be carried on successfully. The
fish would have to be caught in wheels, several of which are operated in the vicinity,
and held until ripe in a retaining channel or pond. This channel has sufficient water
flowing through it, can be easily racked at each end, and seems in every way fairly
well suited for such a purpose. There is a good site for the buildings and sufficient
water is obtainable by gravity from a small creek near by. It is claimed that an
abundance of salmon can be obtained by running the wheels in August and Septem-
ber and that they will be so nearly ripe that retention in the pond or channel will be
necessary for only a short time.

An effort was made in August to transport fish caught in these wheels to the
retaining pond, but it proved a failure. Whether due to lack of care and skill in
making the transfer, or to inherent difficulties, is not certain, but we believe there is no
good reason why salmon may not be safely transferred at this place. We question,
however, whether they would live in the retaining pond until ripe, unless the pond be
rather deep and kept dark.

28 BULLETIN OF THE UNITED STATES FISH COMMISSION.

CELILO.

Mr. A. B. Alexander was directed by the Commissioner to visit Celilo, Oreg., for
the purpose of watching the daily catch of fish taken there in wheels. He arrived
there September 17 and remained till September 23. His report is in substance as
follows:

Only one wheel was in operation September 17. Two others close by were not
running, owing to low water. It was expected to pack 4,000 cases, but by September
23 less than 1,000 cases had been packed. The fall run of salmon was very light in
nearly all parts of the river. The accompanying table will show the daily catch for
one wheel, with their condition, ete.:

Chinook (all caught | Silver (all caught
Steelhead. in wheel). | in wheel).
i}

| & . o | a Pia ae ine | | AES °
| ; ; a elke a le | 2a a] eR ER
| BR) Fae | Sea g/2le/F| |e| ele] #
Dat | 3 a =I) zn ® Ge) sel | tie |e? a | A] 8 Pe
atl steal ene leccie | el berst all Gh | Ale) Se) Bois) ee | e
1) eee ll cee veal ea tee coca ae] =a tera ret ese seth eyes resale res
| | 3 Ss | 8 P|) ee ae eee Pes) conc Foal se Wee Toil celal
| }8 (8/2 | Bl Sls lei 81 ele eis) si sleis| &
| ba /e iS ie 2] She) es fete elelelalz) ae
| linia I CP" les o a elite a} | re o|o a2/8,s5 3
Ye ee ISN ay) ten WSS Mee era et ene ey ees ese te

1896. | | | Ie | |
| Sept.18| 235 | 111 10) 124| 15] 160) 75| 28/19/10| 9| 9) 6| 5) 2) 1| Oo
19 240 130 28} 110 22 180 60 18 8/ 5/10) 6) 8) 6] 4 2 1
21 487 | 121| 29] 366| 98 192 | 295) 53 | 18) 11) 35 | 26) 22| 9 7| 138 &
22| 550| 159] 59 | 391] 164] 350] 200} 20/12) 8| 8] 6|19/11/ 8/ 8/ 6

| | ae!

Total -/1,512 | 521] 126| 991 | 2099 | 882] 630| 119 | 57 | 34] 62 | 47 | 55 | 31 | 21] 24] 15

| | | | | | beee| |

Among the 119 chinooks, 13 were small males, with their milt as fully developed as
that of the large fish. The wheels sometimes take these fish in considerable numbers.
The Indians prize them highly and seldom offer one for sale. Those examined were
quite uniform in size, few weighing over 45 pounds, the minimum being about 3
pounds. In length they vary from 124 to 23 inches. Their color is very dark, the spots
not being visible a distance of 10 feet. On closer inspection all the marks of a large
chinook are visible. Fishermen usually do not class this fish with the chinook, but
think it a different species.

The chinooks taken in the fall of 1896 at Celilo were said to be somewhat larger
than those for several years past, the average weight being 20 pounds; average length,
37 inches; greatest length, 43 inches. The average was about 1 pound larger than the
usual fall run, and 1 pound less than the spring fish.

The steelhead varied in length from 25 to 42 inches, the average weight of those
taken at Celilo being 18 pounds. A few weighed from 35 to 37 pounds. On first
coming from the water the steelhead is very bright-colored, the large specimens
having a bright stripe extending aiong the sides the whole length of the body, vary-
ing from a light pink to a deep bronze. The colors are very pronounced when the fish
is first caught, but grow dim on being exposed to the air.

Silver salmon are from 18 to 30 inches long; their average weight is 7 pounds.

Early in the fall Seufert Bros. operated three wheels, but two of them could not
be run after the river had fallen. Mr. Taffe’s wheel was adapted to low water. An
island lying between Celilo and Tumwater is exposed when the river is low, the

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 29

water on the south side of the river flowing over a rocky ridge forming falls some 12
or 15 feet high. Many salmon find their way to the foot of the falls and endeavor to
ascend them and when the water is about half low many succeed, but when the dis-
tance from the lower to the upper current of water is from 8 to 10 feet thousands of
salmon at times may be seen jumping into the whirlpools at the foot of the falls.
Just above the falls are two wheels, the larger on a rocky point projecting into the river.
When the river is high, many of the salmon pass by this point over that part of the
river where the falls are during low water. The wheel here catches fish only when the
water is high; the other wheel, at the head of a channel, takes fish when the water is low.

The wheels in operation during the fall months are built in places where the water
is deep and the current swift. There being few natural places in the vicinity where
wheels can be successfully operated, long deep channels have been cut through the
voleanic rock and the water led into them. At or near the head of these artificial
channels the wheels are built. The water rushes through the channels with consid-
erable force, offering a strong inducement for salmon to enter. On arriving at the
foot of the falls and finding it impossible to go farther salmon naturally enter the
channel which leads to the wheel. If the water is very clear, the fish will congregate
in large numbers in pools and eddies near the mouth of the channel and at the foot
of the falls, remaining there for days without attempting to go up the channel.
Sometimes several thousand fish will crowd themselves into the mouth of the narrow
channel and remain there for hours at a time. Suddenly, as if by one impulse, they
will dart up channel, and are soon caught up by the revolving wheel and landed on
the platform in the fish-house.

Wheels are built in places near high-water mark where salmon are known to run
in greatest numbers. The migratory habits of all species of salmon visiting the
Columbia River in the vicinity of The Dalles are said to vary from year to year.
This often causes a considerable expenditure of money in wheels frequently followed
by negative results. For several years the main run of salmon may pass certain
points in the river, when suddenly a new route will be selected. This change often
transfers the best fishing-ground from one side of the river to the other, and wheels
which have made large catches one season prove a failure the next. Only those with
long experience in this locality are competent to select good sites for wheels. Several
wheels on the river, built at a cost of from $3,000 to $5,000, have never taken a fish.
This method of fishing requires considerable capital. :

On September 18 and 19 a large number of fish, mostly steelheads, were noticed
to ascend the falls, but, as the river was falling rapidly, in a short time they were
prevented from taking their natural course and were forced either to enter the channel
or remain in the pools or eddies at the foot of the falls. At this time the wheel at
Celilo began to take more fish than it previously had, the daily catch increasing
during the time Mr. Alexander was there. The catch of chinook and silver salmon
was small as compared with the number of steelheads. The last-mentioned species
is always found in greater numbers at this season, the fall run of chinook being
limited in number. Only a few silver salmon are caught here.

This season, owing to the unusual scarcity of fish at Celilo, the cannery there did
not put up any spring fish, the run being very small. The great bulk of salmon was
taken in wheels below the Cascades. Arriving at this point, they seemed inclined to
go no farther up, and the canneries in that locality made large packs. Fish expected

30 BULLETIN OF THE UNITED STATES FISH COMMISSION.

at The Dalles in the spring did not appear until about the close season. A large body
of steelheads passed up the river during that time. Ordinarily the run of steelheads
arrives at The Dalles early in September, but this year most of the run passed by
during the close of the season.

On September 18 the wheel at Celilo took 160 steelheads, 28 chinook, and 6 silver
salmon, and 75 steelheads were taken by Indians with spears and dip nets. Of the
steelheads, 111 were males and 124 females; 10 males and 15 females were in an
advanced stage of development, and would have been ripe in a comparatively short
time; the rest of the catch would not have been ripe until late in the season. Of the
chinooks, 19 were males and 9 females; 10 males and all of the females would have
been ripe by about the first week in October.

During the five days spent at Celilo 1,512 steelheads, 119 chinook, and 55 silver
salmon were examined. Of the steelheads, 991 were females and 521 males; 299
females and 126 males showed considerable signs of development, and would have
been fully ripe by the first week in October.

As the river falls many places are left bare wheresalmon were caught earlier in the
season. Sharp rocks are exposed, small peninsulas formed, and rocky islands appear
where a few months before there was nothing to indicate their presence except an
occasional small whirlpool or riffle. At such places Indians fish with spear or dip net.
Each fall several tribes from various parts of the State camp at Celiloand Tumwater,
remaining there until the season is over. They fish for the canneries principally, but
also lay in a supply of salmon for themselves. In pleasant weather when salmon are
plentiful they do a lucrative business. Sometimes one man catches 35 salmon in a day.
Their fishing is with a dip net and detachable gaff hook and is carried on in swift
water, the men standing on overhanging rocks or shelving places where a foothold can
be had, This is dangerous work, and not a few lives have been lost from time to time.
When the wind is high, one of the strong gusts that frequently sweep the river may
suddenly strike a fisherman unprepared to receive it and blow him into the river.
Fach fisherman who stands in a dangerous place therefore has a rope fastened around
his waist and secured to a rock.

In the whirlpool of water only an occasional fish can be seen and the fisherman
stands and dips blindly. Sometimes a hundred dips or more will be made without a
fish being taken, at other times one will be caught at almost every dip. Where the
water is less turbulent salmon are taken with spears, each spearsman having his par-
ticular standing-place, from which he keeps a vigilant watch, few salmon escaping that
approach within throwing distance.

The Indian is assisted by the women and children of the tribe, who gather the
fish when caught, throwing them into a pile on the rocks and rendering such other
assistance as may be required. As soon as the day’s fishing is over the Indian leaves
the fish to be taken care of by the women, and they are taken from the island to the
mainland, and carried in baskets over a steep, rocky path to the cannery, a distance
of about a third of a mile. When there is a good catch many trips are required to be
made to the cannery. A male Indian seldom does this kind of work; he is quite
willing that it should be performed by the women, and raises no objection when young
girls are pressed into the service by their mothers. Sometimes, however, small boys
lend assistance, but by the time they arrive at the age of 12 they have caught the
spirit of their elders and refuse to do woman’s work.

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 31

When salmon are plentiful an Indian can earn from $5 to $8 a day. The price
varies from year to year, according to the supply and demand for salmon by the can-
neries. This season 2 cents a pound was the standing price for chinooks and steelheads.
Indians consider the chinooks superior to all other salmon, and it is only when plentiful
that they are sold to canneries. No part of this fish is thrown away; the ripe eggs
and even the tail, head, and fins are eaten.

The dwarf chinooks, previously spoken of, not being used by canneries, are greatly
sought after by the women of the tribes, who squat on the floor of the room where the
fish are dressed, patiently waiting for the small fish to be culled out. They sit for
hours shivering in the cold, and consider themselves well paid if each receives one
or two fish. During the fishing season the women and children congregate in the
dressing-room, waiting for a share of the heads, tails,.and fins as they fall from
the butcher’s knife.

The ripe eggs of chinook salmon are considered a great delicacy and are gathered
in large quantities. They are first separated and then dried in the sun, after which
they are prepared in various ways as food.

Steelheads swim near the surface and are more easily caught than the chinooks,
which swim deep. Clear water influences the catch; if the water be clear the catch is
less than when it is muddy.

From September 25 to the 13th of October 2,667 steelheads, 1,402 chinooks, and
2,213 silver salmon were examined at Celilo; 1,010 of the steelheads were males and
1,657 females; 683 of the chinooks were males and 719 females; of the silver salmon,
1,011 were males and 1,202 females; 350 male steelheads and 601 females were in an
advanced stage of ripeness; also 574 male and 528 female chinooks, and 843 male and
1,048 female silver salmon.

The eggs and milt of the salmon caught during October were somewhat further
advanced than that of those examined in September at Celilo. A large part of the
eggs of the chinook on being taken from the fish would immediately separate; this was
also true of many of the silver salmon and steelheads. A number of the latter showed
no signs of development, but many were well advanced and some about ready to spawn,
It would seem that the spawning season of the steelhead extends over a greater period
of time than that of other species. A careful examination carried on in other streams
would no doubt throw more light on the subject.

In the fall, after the close season, a seining “ gang” of twelve to fifteen men goes
into camp on a long pebbly beach, about a mile below the fishing station at Tumwater.
This season two seines were operated on the Oregon side of the river and one on the
Washington side. At times fish will be found plentiful on one side of the river and few
on the other, when suddenly, without apparent cause, they shift to the opposite side.

During the fishing season seining is carried on with little or no interruption, hauls
being made in quick succession all through the day. The seines are set from a flat-
bottom boat, one end of the seine rope being held on shore by horses while the net is
being thrown out. As soon as the net is set the seine rope on the lower end is picked
up by other horses and the hauling-in commences. It would be impossible to land the
net by hand, so swift is the current, and frequently the united strength of four horses
is barely able to land it. At each haul the fish caught are loaded into wagons and
taken to the fishing station, where they are weighed, washed, and put into cars. If
they are to be canned they are thrown into the car in bulk; if shipped east, they are

32 BULLETIN OF THE UNITED STATES FISH COMMISSION.

packed in boxes and iced. In the early part of the fall season the demand for fresh
steelheads is large, and there is more profit in shipping fish east than in canning them.

The only place noticed near Tumwater where salmon could be held for spawning
purposes is near the seining-ground on the Washington side of the river. Here a
channel has been cut, but owing to the small number of fish that entered it the wheel
built near its head and also the channel have for several years been abandoned. On
the lower end of the channel is a small indentation where an eddy is formed; a dam
could be built here and also one across the mouth of the upper end of the channel,
with iron screens at either end. By covering the top of the channel, either the whole
or a part of it, an inclosure would be formed which would hold a large number of
salmon. It seems reasonable to suppose that salmon would live much longer in an
inciosure of this kind than in boxes, as they would be crowded far less and have a
better circulation of water. A swift current or comparatively still water could be had
by putting in a gate at the upperend of the channel. The cost would be considerably
more than keeping salmon alive in cars or boxes, but it is thought that the results
would be more satisfactory.

SEUFERT’S TUMWATER.

On the Washington side of the river, opposite Seufert Brothers’ cannery, is an
excellent seining-ground, where salmon are said to be caught in large numbers.
Within a few rods is a narrow channel of the river suitable in every way for a retain-
ing pond. The hatchery building could be placed upon ground quite close to the
retaining channel and where a supply of suitable water can be led by gravity. Seining
is carried on principally for silver salmon and steelheads; many chinooks are also
caught, but they are so near the spawning condition that they are not used for can-
ning. The probabilities that fish could be secured here and retained for a short time
until ripe are greater than at the Cascades, and the expense would be very slight.

JOHN DAY RIVER.

This river was examined at its mouth September 2. It is there a stream of
moderate size, quite shallow, and very muddy. The banks and bottom for some
distance above the mouth are of mud or sand. Farther upstream the water is, of
course, much clearer. Salmon could doubtless ascend the John Day River, but it
is not certain that they do so in any numbers, and none was observed. Interviews
with persons living on or familiar with the upper course of the stream failed to elicit
any evidence that chinook salmon are now found there. It is said that a good many
salmon formerly entered this river and spawned in the spring, but these are evidently
steelheads and not chinooks. A few years ago a wheel was operated in the mouth of
the John Day, but it has not been run for two or three years.

DES CHUTES RIVER.

The Des Chutes River has its source on the eastern slopes of the Cascade
Mountains and elsewhere in Crook County, Oregon, flows nearly north and joins the
Columbia about midway between the mouth of the John Day River and The Dalles.
It has been generally regarded as the best salmon-spawning stream in the lower
Columbia basin, and it was reported that large numbers of salmon can be seen at any
time in the fall in this river at Sherar’s Bridge, 30 miles from The Dalles or about 40
miles above the mouth of the river. This place was therefore visited from August
29 until September 1,

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 33

Throughout the lower portion of its course this river has cut a narrow, deep
channel in the lava beds. The canyon’s walls are so abrupt that it is impossible to
reach the river in many places. At Sherar’s Bridge the canyon widens out a little and
the river here has a fall of 8 or 10 feet. Below the falls the river is compressed into
a very narrow gorge, through which the water rushes with much fury. At the bridge
the water is said to be 86 feet deep, which is about equal to the river’s width at that
point. Immediately above the falls the river widens out greatly and is very shallow
for some distance.

Until within the last six or eight years “salmon” were seen here every spring in
considerable numbers. Mr. J. H. Sherar, who has lived at these falls for many years,
says “salmon trout” were present throughout the year, but were not abundant except
in the spring, and that they spawned in the spring; they had no difficulty in ascending
the falls, and his understanding has always been that many of them went far toward
the headwaters to spawn. Formerly Mr. Sherar caught large numbers of these
salmon trout, but he has seen but very few for four or five years.

Several persons who live on the upper course of the Des Chutes, report that for-
merly a good many ‘salmon trout” were seen there in the spring, which was their
spawning time, but that few if any have been noticed for several years. It would
appear that the chinook salmon has never run into the Des Chutes in large numbers
and that few enter it now; that the fish which at one time was rather abundant
is the salmon trout or steelhead, Salmo gairdneri. The spawning-beds in this river
are too remote from the railroad to be available for fish-cultural operations.

HOOD RIVER.

This stream is fed chiefly by the snows and glaciers of Mount Hood and flows into
the Columbia at Hood River Station. It does not appear to ever have been frequented
by chinook salmon in large numbers. The people at Hood River Station do not think
that many salmon enter this river, and persons living several miles up the river say that
salmon are seen there but rarely.

BIG WHITE SALMON RIVER.

This river was examined August 6, and again on September 4. It is a stream of
considerable size, having its sources on the western and southern slopes of Mount
Adams. It flows in a general southerly course and empties into the Columbia nearly
opposite Hood River Station. It is, in its lower course at least, a rough, turbulent
stream, full of rapids and large bowlders, and flowing between high, and, in some places,
precipitous, banks, so that it is difficult to reach the edge of the stream. At the
mouth of the river the canyon widens out and there is a little level land along the
stream. In August all of this was flooded by back water from the Columbia, but in
September it was dry. No salmon were seen on our first visit, but on September 4
quite a number were observed jumping in the Columbia about the mouth of the tribu-
tary stream. On that date Indians had established two camps at the mouth of the
river for the purpose of taking salmon during the season. They go up the stream from
a quarter of a mile to a mile and capture the salmon by means of gaff-hooks on the
ends of long poles. Only four salmon were taken during an afternoon. While the
conditions are not favorable for operating a hatchery on this river, a good many eggs
could probably be obtained and carried to any station which may be established on

the lower Columbia.
F. C. B. 1897—3

34 BULLETIN OF THE UNITED STATES FISH COMMISSION.

LITTLE WHITE SALMON RIVER.

This stream enters the Columbia about 6 or 8 miles below the mouth of Big White
Salmon River. It was examined September 5 and 7, and the prospects for successful
salmon-cultural operations appeared so good that a site at the mouth of the river was
selected and preparations begun at once for equipping the station. Salmon were
already in the stream in large numbers and spawning had begun. By the time the
station was ready to begin operations the height of the spawning season had arrived,
but more than 2,000,000 eggs were taken, and it is believed by Mr. William P. Sauer-
hoff, superintendent of the station, that 10,000,000 eggs could have been obtained if
operations had begun earlier in the season. The success attending the operation of
this stationis highly gratifying. Up tothe time of writing these lines (October 30, 1897)
the total number of eggs taken at the Little White Salmon Station this season is over
12,500,000.

TANNER AND EAGLE CREEKS.

These are two small streams flowing into the Columbia at Bonneville,Oregon. The
smaller is Tanner Creek, whose mouth is only a few rods west of the railroad station.
It was examined September 7 by Messrs. Evermann and Hubbard, who found about
300 chinook salmon in the stream within a mile of its mouth. Most of the fish were
lying quietly in deep holes, but some were in shallow water and evidently spawning.

The mouth of Eagle Creek is about half a mile east of Bonneville. This stream is
considerably larger than, TannereCreek. Its bed is exceedingly rough and rocky, and
only with much difficulty can salmon ascend it. A large bunch of salmon were lying
in the deep hole under the railroad bridge and a few others were seen a short distance
farther up the creek. Though these creeks are small, from 1,000,000 to 2,000,000 eggs
could probably be taken there aunually and carried to the Little White Salmon
Station.

Across the river from Bonneville, and only a short distance below Little White
Salmon River, are several streams into which salmon are said to run. Among these
are Little Wind River, Wind River, Smith Creek, Rock Creek, Hamilton Creek, and
Hardy Creek. Some of these are unsuitable for salmon on account of large quantities
of wood which are floated down, and bars sometimes form across the mouths of others,
especially Hamilton and Hardy creeks, so that salmon are not able to enter; but
usually a good many salmon spawn in these creeks, and doubtless a great many eggs
could be collected from- them. Their proximity to the Little White Salmon will make
it easy to carry the eggs to that station.

PUGET SOUND.

In connection with the work of the joint commission appointed to investigate
the fisheries of the contiguous waters of the United States and Canada, certain
investigations were made by Messrs. Alexander and Cox concerning the movements
of the sockeye or blueback salmon in Puget Sound and its tributary streams. The
information obtained has been furnished to the joint commission and has been used
in its report. The movements of the sockeye in the north end of the sound were
studied, particularly among the San Juan Islands, chiefly to determine whether many
of that species run into streams flowing into the sound from the State of Washington.

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 35

The principal rivers examined were the Nooksack, Skagit, Stillaqualmish, and
Snohomish. The time was too short to permit an exhaustive study of the questions
involved, but enough was done to show that most of the rivers flowing into Puget
Sound from the east are salmon streams of considerable importance. This is particu-
larly true of the Skagit. The sockeye runs into this river in large numbers, and there
are important spawning-beds of this species in its headwaters. The State of Wash-
ington has recently established, and is now successfully operating, a hatchery for this
species on Baker Lake, one of the upper tributaries of the Skagit.

The height of the run of sockeye salmon in 1896 at the San Juan Islands was
from August 20 to 28. The run was regarded as unusually large in view of the fact
that this was an off year. Nothing definite was learned regarding the occurrence of
the sockeye in any streams except the Skagit, nor whether it is found in any of the
upper tributaries of the Skagit except the Baker. The Baker River has a lake in its
upper course which can be reached by the sockeye.

So far as is yet known, this species spawns only in such streams as have lakes at
their headwaters, and the spawning-beds are found only in the streams tributary to
the lakes or in the lakes themselves. Several streams tributary to the northern part
of Puget Sound contain lakes in their headwaters, and it would prove interesting and
valuable work to determine accurately the distribution of the sockeye in this region.

All the other species of west coast salmon are found here in large numbers. The
chinook first appear, then the silver salmon and steelhead, and still later the humpback
and the dog salmons. The Nooksack is the principal river frequented by these species.
The silver, hampback, and dog salmons are said to ascend both forks of this river,

The actual or even relative abundance of the different kinds of salmon and their
distribution in the rivers of western Washington is not well understood, and scarcely
anything is known regarding them after they enter the rivers.

LAKE WASHINGTON.

In February, 1889, the United States Fish Commission planted 375,000 fry of the
common whitefish (Coregonus clupeiformis) in Lake Washington at Seattle. In June,
1896, Mr. A. B. Alexander visited this lake and began a series of observations for
the purpose of determining the results of this planting of fish. After a few observa-
tions he was called away to other work, but returned and again took up the inquiry
November 23, continuing it to January 10 following.

General description of Lake Washington.—This lake is situated in King County,
Washington. Its greatest length, which lies nearly due north and south, is 185 miles,
and its greatest width, from Mercer Slough across Mercer Island to the west shore, is
about 43 miles. The greatest width of open water is about half a mile north of Mercer
Island and is 34 miles. The narrowest channel lies between Mercer Island and a
point on the east shore about ? of a mile north of Mercer Slough. The city of Seattle
is situated upon the west shore about equally distant from the north and south ends
of the lake. There are numerous bays and indentations, which give the lake a total
shore-line of about 58 miles. The lake has an elevation of 34 feet above low tide at
Seattle. The principal outlet is Black River, at the extreme southern end of the lake.
This river is 24 miles long, and unites with White River to form Duwamish River,
which flows into Elliott or Duwamish: Bay a short distance south of Seattle. About
a half mile below the lake Black River receives the Cedar River from the east,

36 BULLETIN OF THE UNITED STATES FISH COMMISSION.

A few years ago a channel was cut from Union Bay through Lake Union to the
sound. Considerable water flows through this channel and small boats are able to
pass through it.

At the north end Lake Washington receives Squak Slough, which is the outlet of
Lake Sammamish. This slough runs in a tortuous channel through low, marshy
ground, its width varying according to the height of water in the lakes which it con-
nects. When the water is very high the slough overflows and covers the low land on
either side.

During heavy rain the Duwamish River is unable to carry off readily all the
water brought down by the White, Cedar, and other streams, and as a result Black
River becomes, for the time being, an inlet instead of an outlet of Lake Washington.
The lake therefore rises perceptibly, the inflowing water from Lake Sammamish is
held back, and Squak Slough overflows the adjacent lowlands.

Lake Washington is a beautiful body of water, surrounded by high hills from
which a fine view of the surrounding country may be had. The Cascade Mountains
are plainly seen to the eastward and on the south Mount Ranier comes into
conspicuous view. In many places steep banks make down to the water’s edge,
indicating deep water. Nearly all the hills are covered with a dense growth of
trees, except where clearings have been made for homes and settlements. The shore
line in nearly all parts is fringed with a dense undergrowth of brush and small trees;
tule grass is found at every low point and slight indentation. Small trees along the
shore interfere with the hauling of seines, and only limited collections can be made.
In November and December a large amount of decayed vegetation covers portions of the
shore. In summer various forms of minute animal and plant life inhabit the water.

Seine-hauling grounds.—Only in a few places along the shore of the entire lake is
the bottom sufficiently free from snags, fallen trees, and other material to permit the
successful hauling of nets. On the west side of the lake, about 2 miles above Union
Bay, is a long stretch of sandy shore mixed with fine gravel; in June, 1896, several
attempts were made to haul a seine here, but on account of the many obstacles no
fish were taken. On the east side, at Hunter Point, is a comparatively good ground,
but tule grass grows in considerable quantities, though by wading out and bending it
down where the seine is to be landed fairly good results can be obtained. Several
hauls were made here and a number of fish taken. In fall and winter there is a
great scarcity of life here, as several trials demonstrated. From Hunter Point south-
ward for about 24 miles the shore makes down to the lake with a gradual slope, but
owing to the great amount of vegetation that extends in most places into the water, it
is almost impossible to use a seine. This also may be said of Mercer Slough and the
shore farther south. On the east side of Mercer Island the bank is usually quite steep ;
on its west side the land is more sloping. The water all round the island is deep. In
the extreme southern portion of the lake, in the vicinity of Renton and Black River,
the land is marshy and filled with small sloughs. The bottom is of soft mud.

On the north and west sides of Lake Washington, from Webster Point to Sand
Point, are scattering sandy patches, but only in a few places can a seine be hauled,
owing to a thick growth of scrubby trees at the water’s edge. High banks in most
places overhang the shore line. Northward from these points the land is more sloping
and in some places comparatively level. Rocky and gravel beaches crop out from the
banks in many places and in a number of indentations the water makes off shallow

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 37

from the shore a distance of 50 feet or more. The bottom here is very muddy and
covered with tule. The shore on this side of the lake is thickly strewn with logs,
slabs, and small floating débris. Few rocks are noticeable, and none at any great
distance from shore.

The sandy and gravelly beaches referred to on the west side of the lake below
Sand Point are usually narrow; only in one place observed is the beach over 10 feet
wide; here the width from the water’s edge to the bluff is about 30 feet. The clear
ground indicates a smooth bottom, but stumps of trees and broken limbs prevent
seine collecting. These beaches are formed by wave action, the water having washed
away the base of the bluffs, leaving a deposit of sand and gravel. In many places,
particularly on the west side of the lake, bushes and small trees have taken root in the
sand and gravel deposit. Only in a few instances does the sand and gravel formation
extend any great distance below the water’s edge; 30 or 40 feet from the shore is soft
muddy bottom. Asis indicated by the high perpendicular bluffs, the beaches make off
at a sharp angle, deep water being found a short distance from the shore. As a
whole, the shore of Lake Washington is not well adapted to collecting with a seine.

Depth and character of bottom.—Soundings were taken at different times and in
various parts of Lake Washington. It has a uniform depth in the center, ranging
from 138 to 222 feet. The character of the bottom is generally muddy. Off the mouth
of Union Bay and near the boathouse at Madison street sandy bottom was found. Off
and in the vicinity of Sand Point, 34 miles farther north, the bottom is composed of
sand and gravel. This kind of bottom does not extend very far from the shore, from
500 to 800 feet being about the limit. For nearly 2 miles south of Sand Point there
are a number of sandy places along the shore, mostly covered with an undergrowth
of bushes. In all other parts of the lake examined the bottom is chiefly of soft mud.
On the east side of the lake, south of Hunt Point, the water is quite deep close to
the shore; this is also true of the west side, south of Sand Point, and in many other
parts. A depth of from 500 to 1,800 feet was reported off the northern end of Mercer
Island and in the vicinity of Hunt Point. Comparatively deep water was found oft
these points, as will be seen by referring to the map, but no such depths as had been
reported. About 1 mile along the shore above Kirkland a plateau from 60 to 100 feet
wide, covered with vegetation, makes off from the shore. The water varies in depth
from 10 to 20 feet, but suddenly deepens into 60 and 70 feet when the edge is reached.

Temperature of water.—It will be noticed in the table of temperatures that near
the shore on the east side of Mercer Island the temperature of the water at the surface
varies considerably from that taken a short distance toward the middle of the channel.
A difference of 25° is recorded, the bottom temperature in both soundings being the
same. In no other part of the lake was such a marked difference found. Several
tests of the water were made within a radius of 100 feet or more, all with the same
result. A few hundred feet from this position, to the north, south, or east, 47° was
found at the surface. It is said that thermal springs occur in this vicinity in a
number of places. The temperature of the water at the bottom, however, does not
indicate their presence. The lowest temperature found in the main body of the lake
was 444°, and the highest 483°; the first in depths of 180 and 190 teet. In Union Bay,
in 14 feet of water, a temperature of 41° was recorded both at the surface and the
bottom. This low temperature was due, no doubt, to the cold weather, which at this
time was quite severe. At no other place was such a low temperature found.

38

The temperature of the air seemed to have no perceptible daily effect on the
water in deep portions of the lake. From November 23 to 30 the air varied but little,

from 20° to 29° being about the minimum and maximum height.

BULLETIN

from 45° to 48° was the approximate daily temperature.

In depths over 150 feet the air ceased to have an effect on the water much below
In a few places 50 feet below the surface was reached before
In depths less than 100 the temperature was

40 feet from the surface.
the temperature remained constant.
found constant about 30 feet below the surface.

It will be noticed that in a number of places in depths ranging from 40 to over
200 feet the temperature botlat the surface and bottom is the same. Why this should
be, when only a short distance away the difference between the surface and bottom is

a degree or more, is not easily explained.

Many of the temperatures and soundings taken are recorded in the following table:

OF THE UNITED STATES FISH COMMISSION.

Soundings and temperatures taken in Lake Washington.

CMa ee || peel ie
a8 |e | B22) ee
ef \)ss 28 | as
Date. 3 |S2)|s6 Position. Date. 2 5 5 2s Position.
2 Erm | eo = a | &
BE Ete ibe Bee es
A |a a fe) tebe |S!
Feet.| °F. | OF. l Feet.| °F. | °F,
Noy. 23 35 483 483 300 feet from shore, near | Dec. 19*) 5 493! 46 | Close to shore east side of
Madison-street boathouse. Mercer Island.
24 14 5 44{| In various parts of Union | 19} 58 47 46 | 300 feet from shore and
24 10) 45 45 Bay, which is a part of | | near last sounding:
\| Lake Washington. | 19| 67} 47 46 alfway across channel
25 98 48 48 | One-fourth mile east of boat- between Mercer Island
house. and east shore.
25 | 156 48 47 | One-fourth mile east of last 19 55 47 46 | Near the east side of lake.
| | station. 19 12 46 45 | Close to shore.
25| 162/ 48 47 | Middle of lake. 20 70 46 46 | Half a mile north of last
25 | 150) 48 47 | Two-thirds across lake. | sounding.
25) 144 48 47) One-eighth mile from east 20 87 46 46 | One-eighth mile west from
| side of lake. | | last sounding.
25 72) 48 48 | 300 feet from shore. | 20) 1153, 46 47 | Halfway across east end
27 14 41 41 In Union Bay; cold weather, | Mercer Island to east
| air 28° and 30° ] | _ side of lake.
27 | 98 47 44 | Off mouth Union Bay. 20) 174 463 46 | Halfway between Daphne
28 | 150 URN ea Do. Point and north end Mer-
28 | 152 47, 47) Do. | cer Island.
Dec. 1 96 47 47 Do. 23) 204 46; 453 Half a mile east Webster
“1/] 168; 47 7 | 14 mile SSE. from Madison- | Point.
street boathouse. 23 | 219 464 45} Half a mile north last
1} 90 47 47 | 13 mile east from Yesler- sounding.
street boathouse. 23} 201 464, 453 Half «mile northward last
1) 144))) 47, 46 | Two-thirdsacross lake from | | sounding, { mile from
| j pumping station to Mer- } | shore.
| cer Island. 23 66 46 46 Third of mile north and
1| 156 7\ 46) One-halfmileoffwaterworks. | | close to shore.
18 | 210 464, 46 | One mile SE. from Madison- 23 584, 464) 464 About 500 feet from Sand
| street boathouse. | Point south near shore.
18 | 204 46 46 | One-fourth mile SE. from 23} 105} 47 46 About 200 feet from Sand
| | Madison-street boathouse. | | Point NE.
18 | 218 463 46 | 1} mile SE. from Madison- 23 | 181 464 45) Less than } mile NE. from
| | street boathouse. | | | last sounding.
18 | 204) 46, 46 One-third of a mile SW. of 23 | 168 464 45 Less than $ mile in same
| | _ Daphne Point. | direction.
18 | 150} 46 46 One-eighth of a mile SW. 23 | 180 46} 44: About the same distance
of Daphne Point. | farther on.
18| 156) 46 46 One-third mile off Meyden- 23 | 190 463) 444 Less than 4 mile NE. last
| bauer Bay. | | station. a
19| 114 46, 46 Near middle of Meyden- 23) 178 463 45) Last een sounding on
bauer Bay. this line.
19 69 46 | 46 | South side of bay. (Other 23 72 46 46 | Near snpoatee shore from
soundings taken in bay; Sand Point
| same temperature. ) 23 96 46 46 | About1,500 feet from shore.
19 12 473, 47 | Short distance south of 23 | 129 46 46 | About halfway from last
| _Meydenbauer Bay. station to Kirkland.
19 40 47 47 | One-third mile farther

south near shore.

a Said to be thermal springs in near vicinity, which may account for the change of temperature.

During December

Bull. U.S. F.C, 1897. (To face page 38.) PLATE 1.

SEATTLE.

Map of Lake Washington, Wash., show-
ing soundings, in feet, taken by the
U. S. Fish Commission, in November
and December, 1896.

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 39

During the winter months short periods of cold weather sweep over this part of
the Northwest, but not sufficiently severe for ice to form in the main portion of Lake
Washington. Occasionally skim ice appears on the margin of the lake in sheltered
places covering small areas.

Trials with gill nets and other apparatus—On November 25 the first trial with
gill nets for whitefish was made in 35 feet of water, 500 feet from the shore, at a point
where it is said that an example of this species had been taken. Two nets were set,
and remained down 48 hours. They were visited on three different occasions during
that time, but no fish were found in them. The reason for setting nets in such
shallow water for whitefish, was in compliance with a request made by the person who
claimed to have caught a specimen on the above-mentioned ground in the fall of 1892.
It was taken on a hook, which throws considerable doubt as to its being the true
whitefish. So far as we have been able to learn no other specimen has since been seen.

On November 24 three more gill nets were set about 14 miles from first position
on the south side of Union Bay. Here also whitefish had been reported. Depth of
water, 15 feet; temperature of water at bottom and at surface, 45°; bottom muddy.
These nets remained in the water three days, and were visited each morning. No fish
were found in them.

On November 26 three nets were set at the head of Union Bay, in 10 feet of water,
close to the shore, where fish are said to be plentiful in summer; also two other nets
near outlet or portage, about three-quarters of a mile farther south. In this place
anglers have good fishing, and it was thought that if fish were running in any consid-
erable numbers the nets would be likely to take samples of different species. The
nets set at the head of the bay were taken up the following morning, nothing being
found in them. Those set near the outlet remained down two days, and were visited
each morning. The second morning one cut-throat trout and one chub were taken
fromthem. During the time the nets were down a number of cut-throat trout, chub,
and suckers had been taken on hook and line. Sounding and collecting with surface
tow net had also been carried on whenever the weather permitted.

On November 27 three gill nets were set near the bottom, off the mouth of Union
Bay, in 18 feet of water, bottom hard sand, where they remained two days. Nothing
was found in them when taken up. It is reported that large numbers of fish pass over
this ground, and it was thought that our efforts would be rewarded by a few specimens.
The result of this trial indicates that at this season fish do not move in great numbers.
The same nets were set in 96 feet of water about a quarter of a mile farther off shore,
where the bottom temperature was 5° warmer than at last station. The nets were,
as before, set near the bottom, but without result. Had whitefish or other species
been traveling over this ground a few would have been captured.

A small creek that flows into the head or western part of Union Bay was next
investigated. This creek is said to abound with trout in summer and with redfish in
early fall. The water in the upper part of the creek flows with considerable velocity
over a gravelly bottom. It is from 4 to 5 feet wide in most places, and about 7 or 8
inches deep; there are a few pools whose depth is nearly 2 feet. The creek is largely
fed by seepage from the surrounding hills. At the time of our visit the weather was
cold and the ground covered with snow, and in many places ice had formed. The
stream empties into a marshy lagoon where it finds its way, through several smaller
ones, into Union Bay.

40) BULLETIN OF THE UNITED STATES FISH COMMISSION.

Repeated trials with a small collecting seine at the mouth of the creek were made,
but nothing was taken in it. Only one fish, thought to be a redfish, was seen, and it
escaped under a fallen tree. Unsuccessful hauls were also made from a half to three-
quarters of a mile upstream. Hook-and-line fishing was also tried, with the same
result, salmon eggs being used for bait. At the mouth of the creek a salamander was
found, and this was the only form of life seen, except the supposed redfish.

On December 11 gill nets were set in a number of places, from a half to three-
quarters of a mile apart, in Squak Slough at tne head of Lake Washington,in 9 and 10
feetof water. They were visited twice in 48 hours, nothing being foundinthem. The
bottom temperature in the places where they were set was 444°, surface 453°. During
the entire investigation the water in the slough was unusually high, and this prob-
ably had a great deal to do with the scarcity of fish here, for at such times, instead of
following the main channel, fish would naturally scatter over a large area, thereby
greatly lessening the chances of taking them. ‘Troll fishing was tried, but with unsat-
isfactory results.

Nearly allof December 18 was spent in sounding and taking water temperatures.
In the afternoon four nets were set on the north side of Meydenbauer Bay; depth, 102
feet; bottom and surface temperature, 46°. Nothing being in the nets the following
morning they were taken up and set on the north side of Mercer Slough, nearly 3 miles
farther down the lake. A trial for 48 hours resulted in the capture of one female
cut-throat trout, the eggs of which were approaching a state of ripeness. Two other
specimens subsequently taken showed the same amount of development. Several
creeks on the west side of the lake were investigated but no fish were found.

Whitefish.—There is no satisfactory evidence of whitefish ever having been caught
in Lake Washington, though several persons claim to have seen them. One person
says in the fall of 1892 he caught a specimen, but no accurate description was given
as to its shape or color, and Prof. O. B. Johnson and State Fish Commissioner James
Crawtord doubt whether it was the genuine whitefish, and believe that the whitefish
planted in the lake no longer exist. The investigation strengthens this opinion. The
Columbia chub (Mylocheilus caurinus) is a common species in the lakes of this region,
and is by many people known as “ whitefish.” It is not improbable that all the
so-called “ whitefish” which have been reported from Lake Washington were really
this species of minnow. Williamson’s whitefish (Coregonus williamsoni) doubtless
occurs here, but no specimens were seen.

Other fishes of Lake Washington.—The various species of fishes collected during
these investigations, as well as all other species known to occur in Lake Washington,
are listed in another part of this report. The principal species are the eut-throat
trout, Columbia River chub, squawfish, Columbia River sucker, a blob, two or three
other species of Cyprinida, and the redfish. Salmon are said to enter the lake
through Black River early in the fall, but none was seen. They are probably the
large form of the redfish or sockeye (Oncorhynchus nerka). Redfish are said to run up
into shallow places during the latter part of October and a part of November, but
repeated trials resulted in collecting only a few specimens. Those familiar with the
different runs of fish on the lake attribute the scarcity of redfish this season to high
water. In Squak Slough, where they are said to run in great numbers, not a specimen
was Seen or taken, though before the water began to rise in the slough a considerable
body of these fish was reported as passing through. Black bass are occasionally

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 41

taken in Lake Washington, and they are at times quite plentiful in Lake Union, which
supplies all the black bass to the Seattle market. The yellow perch (Perea flavescens)
has also become well established in this lake.

LAKE SAMMAMISH, WASHINGTON.

While engaged in carrying on observations upon Lake Washington Mr. Alexander
made a trip, December 31 to Jauuary 6, to Lake Sammamish, a short distance to the
eastward. He gives the following account of this lake and the investigations:

This body of water lies parallel to Lake Washington, the two lakes being sepa-
rated by a hilly strip of country, the most narrow part being 35 miles across. The
greatest distance between the lakes is 8 miles.

Lake Sammamish is 388 feet above low-water mark at Seattle and 4 feet above
Lake Washington. The land separating the southern portion of the bodies of water
might be termed mountainous; many of the hills rise from 1,000 to 1,500 feet from
their base, and are covered mostly with a dense growth of trees. A road cut through
the forest from the west shore of Lake Sammamish to the east side of Lake Washing-
ton is traveled considerably during the summer months. A boat connecting at the
end of the road carries passengers across Lake Washington and lands them in Seattle
much sooner than if they traveled by rail.

The length of Lake Sammamish is 8 miles; its greatest width (about halfway
between its north and south points) is 13 miles. lts narrowest part is near the
northern end, a little over half a mile from shore to shore. From Inglewood, in the
northern part, to Monohan on the south it has a nearly uniform width of about 14 miles.
It lies nearly in a north-and-south direction; a line being drawn north from the south
end would strike 1 mile east of its most northern part. The Seattle and Lake Shore
Railroad strikes the lake at Adelaide, and follows the east shore the entire length.
Three small towns are situated near the shore on the east side. The most important
is Monohan, near the south end. On the west side a number of settlers have made
considerable clearings. Opposite Monohan is a huge pile of sawdust, which extends
for a considerable distance into the lake, and when the water is rough small chunks
of it are washed away. It being water-soaked, most of it that falls into the lake
immediately sinks. The sawdust coming from the mill at Monohan is burned.

Topography of the shore—The shore in most parts of Lake Sammamish is com-
posed chiefly of sand and gravel, forming numerous beaches. These beaches, when
the water is at its normal height, are exposed. At the time of our visit to the lake
(December 31 to January 6) the water was about 4 feet higher than in summer; it had
then fallen 2 feet since the middle of December. Judging from the condition of the
lake at high water, the beaches when exposed must be from 8 to 12 feet wide, measur-
ing from the water’s edge to the bank. In most places they extend offshore at a
sharpangle. Except in afew places, where the sandy and gravelly beaches appear, the
bottom makes off gradually and its character is either hard or soft mud. The south-
ern shore of the lake is skirted by low marsh land, which extends from the southwest
to the southeast side. Here several small creeks empty into the lake, which, no doubt,
have contributed largely to the marsh land.

The hills bordering the southern portion of the lake descend with a more gradual
slope than those on the east and west sides. The land in the immediate vicinity of
the lake is not more than 4 or 5 feet higher than the level of the water, gradually

42 BULLETIN OF THE UNITED STATES FISH COMMISSION.

rising for a distance of about a mile before the foothills are reached. Many portions
of this slope have been cleared. Off the points in this part of the lake, from 30 to 40
feet from the shore, thick clusters of tule are submerged in water from 7 to 10 feet
deep. Later in the season a large portion of the ground in which this grass grows is
exposed. Soundings were made along the outer edges of these strips and points
without finding any indications of hard bottom.

On the east and west sides of the lake the steep hills rise within a short distance
from the water; in many places projecting points and bluffs rise very abruptly, with
little or no shore line at their base. The bottom close to these projections is hard,
being composed of a mixture of sand, coarse gravel, and small stones. The hilly land
is covered with a dense growth of trees, extending to the water’s edge, and at times
when the water is high it 1s very difficult to walk any distance along the shore
without swinging from one bough of a tree to another. When the water is low or at
its average height the sand and gravel beaches here are bare.

At the foot of the lake, on the northeast side near Squak Slough, a low point
of land extends for a considerable distance southward. From this point south, a
distance of nearly 2 miles, the shore is composed of sand and gravel beaches. On
the west side and north from where the last line of soundings was begun the hills
gradually descend to the shore, and in a number of places toward the foot of the
lake the land is quite level. The shore line along the northwest and northeast portions
of the lake is covered with logs, slabs, and pieces of timber, refuse material that comes
from the shingle mill situated near the north end and the sawmill near the south end.
The greater part of this material is found in the northeast portion of the lake, being
forced there by the southwest winds.

On the west side of the lake, near its southern end, is a submerged forest, about
half a mile long north and south, extending nearly a third of the distance across
toward the east side. The northern end of the forest commences near the pile of
sawdust already mentioned. It runs parallel to the shore and is separated from it
by a distance of about 600 feet. During the summer months it is said that the tops
of many of the trees project above the water from 1 to 2 feet. The lake then is much
clearer than in winter, and on a bright day, when the water is smooth, a large portion
of the forest, according to the statement of people living at the lake, is plainly visible.
Owing to the muddy condition of the water, caused by the heavy rains, we were
unable to find any of the trees. Many theories are advanced by people living near
the lake as to how the forest came to be in its present position. The most common
theory is that at some remote period a gigantic landslide occurred and the displaced
material found aresting-place at the bottom of thelake. Considering that the nearest
hills where a landslide of such magnitude could have started are situated nearly 2
miles from the lake with comparatively level country lying between, this theory is not
altogether tenable. A large tract of land becoming detached from the side of a moun-
tain and carried along rapidly by its own weight a distance of a mile or more from its
starting point would by the time it reached its destination be a huge mass of trees,
earth, and stones. The trees would be more or less broken, and not left standing in
an upright position. It is also likely thatif the forest in question had suddenly been
detached from the side of one of the adjacent mountains it would have stopped on
arriving at comparatively level ground, but instead we find it deposited a long distance
from hills of any great size. If, however, such had been the case, and it was forced

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 43

along over level ground, we might expect to find that part of the lake where the land-
slide was deposited more shallow than elsewhere; but, on the contrary, we find a
depth of water varying from 50 to 70 feet, which corresponds with depths found in
other parts of the lake the same distance from the shore. It seems more reasonable
to suppose that at some distant period the ground on which this forest grew, subsided.
This, in a measure, would account for the trees being left standing in their natural
position. Not having seen any portion of the submerged forest, we are not able to
state from personal observation anything concerning it, only stating what has been
given by several reliable parties. Efforts were made to locate some of the trees, but
none of them could be found with the sounding line.

Depth of water and character of bottom.—Lake Sammamish is quite uniform in
depth, varying from 70 to 90 odd feet in the middle and from 35 to 50 feet near the
shore. The deepest sounding was found near the middle of the lake, about halfway
across from Hattie Bell Springs to the west shore. A line of soundings run across
the upper part a short distance below the shingle mill corresponds nearly with the
line across the south end of the lake. In no part of the lake examined are there indi-
cations of plateaus or shelves making off from the shore. Only in a few places was
shallow water found, and this extended but a short distance from the shore. In nearly
all parts of the lake along its margin the bottom drops off suddenly, the water being
deep enough for boats and small steamers to lie alongside and discharge freight or
passengers.

From 80 to 100 feet from the shore is the average distance where the sand and
gravel merge into the hard or soft mud. A number of places, however, were found
where sand and gravel bottom extends from 125 to 300 feet into the lake; one of these
places is on the west side, opposite Hattie Bell Springs, and the other off the wharf
on which the sawmill at Monohan is built.

At the foot of the lake, above the shingle mill, the bottom is very soft and in the
center between the east and west sides the mud is said to be several feet deep, and this
statement is probably not exaggerated, for in running the line of soundings which ends
just below the shingle mill the sounding lead, nearly all the way across, sank into the
soft bottom a foot or more.

Inlets—Numerous small creeks empty into the lake both on the east and west sides,
but the discharge of water is not great even during the rainy season, and in summer
many of them dry up, they being fed by the drainage from the surrounding hills. The
principal inlets are Issaquah and Manner creeks, which enter the lake on the south
end. The main portion of the first creek is 2 miles long and the second about 3 miles,
each having several branches from 4 to 24 miles long. The water in Issaquah Creek
is said to be very clear during the summer and fall months, or was previous to the
coal company using the creek as a dumping-ground for coal screenings, since which
time the water has become muddy. There are a number of coal mines in this region,
and the people complain of the screenings from them being dumped into the creeks,
thereby doing considerable injury to the fishing- grounds.

Outlet.—Squak Slough, or Sammamish River, is the only outlet to the lake. This
at times is not large enough to prevent the water in the lake from rising rapidly, and
when Lake Washington rises at the same time, which is usually the case, and forces
the water from the slough into Lake Sammamish, both lakes rise rapidly. At the

44 BULLETIN OF THE UNITED STATES FISH COMMISSION.

time of visiting Lake Sammamish the water had overflown the channel of the slough
from a quarter to a half mile; its average width for the part visited is less than 100
feet, in some places not over 50 feet. Its length is about 12 miles; its course, as indi-
cated on the map, is very crooked, and it is navigable for small craft only.

Temperatures.—The water temperatures during our investigation varied but little,
and in many parts of the lake there is no difference. In the first line of soundings
run across the lake the surface and bottom temperatures were the same. The weather
at the time was rainy, misty, and cold. In the second line a difference of 1° was
found in two places, each in a depth of 90 feet. At the surface the thermometer
registered 45° and at the bottom 44°. In the third line the depths were from 33 to 92
feet, with no variation in temperature. The fourth line, run the same day, shows in
eight different soundings 45° at the surface and 445° at the bottom. The fifth line,
run across the head of the lake, showed no difference; but in a number of soundings
taken on the west side, near where a small brook empties into the lake, the water was
1° warmer at the surface than at the bottom. The depths here were comparatively
shallow. The last line of soundings run, which was across the foot of the lake, also
showed the water to be 1° warmer at the surface than at the bottom. This difference
may possibly have been because the weather was warm and pleasant, while most of
the time, when the other soundings were taken, it was rainy and foggy and the
temperature of the air from 44° to 50°. On this day, January 6, it registered 60° a
greater part of the day.

Where the water at the surface was warmer than at the bottom it was found that
the warm water extended from 5 to 8 feet below the surface; where the temperature
at the bottom was warmer than at the surface the water was constant from 15 to 18
feet from the bottom. The thermometer was let down at various depths, in most
cases at intervals of 10 feet, no variation in temperature being noticed until the above
distance from the bottom was reached. In a few places, however, within 10 feet of
the bottom the water remained constant.

Condition of water.—In the fall and winter, when the lake is high, the water is con-
siderably discolorea by vegetable matter and sediment washed from the surrounding
hills and brought down by the swollen creeks. This gradually disappears as the rain
ceases, after which the water becomes clear. The fish then bite better than at any
other time, there being less food in the water for them to subsist upon.

Lake Sammamish, like Lake Washington, never freezes over. When the weather
is unusually severe skim ice sometimes forms near the edge in shallow places, but the
deep water does not freeze.

Fishing trials.—From December 31 to January 6 sounding, taking temperatures,
collecting with surface tow net, and fishing with gill nets and troll were carried on.
The weather most of the time was stormy. Fishing with gill nets was unsuccessful.
They were set in various places and in different depths, visited twice each day, and
in the evening set in a different place. Inquiries were made of fishermen as to where
fish would most likely be found, but nothing was caught. A trawl was the only
stationary apparatus used that captured anything, and this caught only one chub and
a blob.

Fishing with troll line was fairly successful, five cut-throat trout being taken.
The weather was unsuitable for this kind of fishing. To meet with good results the
weather should be moderate and much warmer than it was during the time these trials

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 45

were made. Fishermen say that in summer, and at other times when the weather is
pleasant, it is comparatively easy to catch from twenty-five to thirty trout in a day by
trolling.

Trout would bite only when the wind was light and the lake smooth. During the
day, whenever there was little wind stirring, they could be seen jumping or breaking
water in all parts of the lake, but they did not bite as ravenously as was expected,
probably on account of the creeks having brought down considerable quantities of
vegetable and other matter, such as they may feed upon. Fishermen claim that the fish
never bite well when the lake is high. Several good examples, however, were caught
in the few trials made. When fish are at all inclined to bite, a trawl will usually
pick up a variety of species. The trawl set (90-odd hooks) was baited with fresh chub
and meat, and good results were expected of it, but in this we were disappointed.

The trout taken by trolling were caught close to the shore, that being the part of
the lake where the fishermen troll when the water is high. Fish were breaking water
in the middle of the lake, but none was captured. In summer good fishing is to be
had in all parts.

Methods of fishing.—The principal method of taking fish in Lake Sammamish is
by trolling, though good catches are sometimes made with set line and with rod and
line. In summer the lake is quite a resort for sportsmen from Seattle. No commercial
fishing is carried on, all fish caught being locally consumed, except the few taken to
Seattle by people belonging there.

Kinds of fish—The economic species of fish found in Lake Sammamish are the
same as those of Lake Washington, namely, cut-throat trout, ‘‘ grayling,” or redfish,
and salmon. Noexamplesof the last two species were found, they having disappeared
shortly after the water in the lake began to rise in November. About that time it is
said that redfish were plentiful in Squak Slough and in the creeks at the head of the
lake. Salmon run with the redfish, but only scattering individuals are taken. No
information could be obtained as to what kind of salmon enter the lake; but the dog
salmon and an occasional steelhead probably find their way into this body of water.
During the redfish run the neighboring Indians do considerable fishing on Squak
Slough, and lay in their supply for winter. Only one person living near the lake
claims to have caught a whitefish in it, and he was not sure that the specimen was a
genuine whitefish; others who saw the fish say that it was not, but admit that the
fish differed from any previously taken in the lake.

Spawning habits—People living near Lake Sammamish say that the cut-throat
trout spawn in the spring and the condition of most of the specimens indicates that
they are correct in their observations. One taken in Lake Washington contained eggs
well advanced, but this fish would hardly indicate a difference in the spawning season
between the two lakes, as fish can easily pass from one lake to the other. The redfish
probably spawn some time in November, that being the time they run in greatest
numbers. Only a few specimens were taken, and they were not sufficiently developed
to form an estimate as to the month in which they spawn, but people who have
caught and observed the redfish, both in Lake Sammamish and Lake Washington,
believe that the spawning season is from the latter part of October to the middle of
November. To gain a better knowledge as to the habits of this species in this region
some person should be stationed at these lakes early in the fall, before the rainy season
comes on,

46 BULLETIN OF THE UNITED STATES FISH COMMISSION.

It is said that redfish run only for a short time, that they disappear as suddenly
as they come, and few are seen in any part of the lakes after the run is over. Almost
every person with whom we conversed on the subject expressed the opinion that the
redfish, or “‘ grayling,” as they are commonly called, annually come in from the ocean;
but some believe that they never leave the lake, but retire into the deep parts of the
lake as soon as the spawning season is over. Positive knowledge on the subject is
lacking. The few specimens taken in Lake Washington would hardly warrant an
opinion to be formed as to their habits in these lakes. It would be necessary to observe
a considerable number from the time they first appear until they leave.

Surface material.—Collecting with surface tow net was carried on at different
intervals. As was the case on Lake Washington, surface life was found most abun-
dant in the evening; at other times a trial of 15 or 20 minutes would result in a small
quantity of material.

Depth of water, temperature, and character of bottom of Lake Sammamish.

Tem- | Tem Tem Tem-
Depth Character per- per- Depth Character per- per-
Date. of of ature | ature | Date. of of ature | ature
water. bottom. of of | water. bottom. of of
surface. bottom.
1896. Feet. oF. oF, 1897.
Dec. 31 | a@194 | Hard sand.-.-. 45 45 Jan, 2
31 61 | Hard mud .... 45 45 2
31 90 | Softmud...... 45 45 2
CPO) OPA | eee SAGs 45 45 2
Gil} th) Eb -S--G Gee ssense 45 45 2
Sly ace cee openers 45 45 2]
CHUN EPPO Re = te Ss Sosa ad 45 45 2 |
31 Oven === == 45 45 2
31 d mud, 2
Pniateaerers 45 45 2
3h; 13 +} Gravel. -.-- =: 45 45 2
1897. 2
Jan. 1} 632 | Sand.--....-..- 45 45 2 Hard sand .
1 Soft mud. 45 45 2 60 Soft mud. . 45 45
1 i 45 45 2 45 45
1 e200 45 45 2 45 45
1 BIO Okeiem eee 45 44 2 45 45
Ay a0 aero ee Stee 45 44 | 2 45 45
1G) ae S75 p | eeeeedo eee ners 45 45 | 2 | 45 45
We VERA SG OFBERE SSSA 45 45 | 2 45 45
te ee eee Ole eee 45 45 2 45 45
1) VOR NE Reker isa see 45 45 6 46 45
TE lene Ct? SIE Sr (Ne see Sh 45 45 6 46 45 |
1 (AD se So se= 45 45 6 46 45 |
1 Sand and | 6. 46 45
gravel ...... 45 45 6 45 45
1 Ore 45 45 6 46 45
2 Soft mud...... 45 45 6f 46 45
2 0 --- 45 45 6 46 45
2 45 45 6 46 45
2 45 45 6 46 45
2 45 45 6 46 45
2 45 45 6 46 45
2 cacce tec 45 45 6 46 45
2 33 Hard sand..-. 45 45 6 46 45
2 d324 Hardsandand 6 46 454
gravel ...... 45 45 6 104 | Coarse gravel. 46 46
2 834 | Softmnd.....- 45 444
a Abont 20 feet from bridge; first line sounding. dCommencement of fourth line of soundings.
b Second line of soundings 250 feet from sawmill. e Soundings close to a brook.

ce Commencement of third line of soundings. Jf Sixth line of soundings 20 feet from shore.

Bull. U. S. F. C. 1897. (To face page 46.) PLATE 2.

- Adelaide.

< Shingle Mill.

NORTH

»Monohan,
| 2nd line of soundings.
Saw yet,

Lake Sammamish, near Seattle, Wash.,
showing soundings taken by U. S.
Fish Commission, in December, 1896,
and January, 1897.

Sch line of soundings.

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. AT

LAKE UNION, WASHINGTON,

The following notes on Lake Union have been furnished by Mr. Alexander:

This body ot water lies close to the city of Seattle. Its length, measured in a
straight line from north to south, is 14 miles; width, halfway between the north and
south ends, three-quarters of a mile. Branching out from the head of the lake are
two arms, one leading to the northeast and the other to the northwest. Measured
‘from the foot of the northeast arm the lake has a length of 34 miles.

Lake Union has a quite uniform depth in the center; the greatest depth, 48 feet,
is found in nearly all parts of the main body a quarter of a mile from the shore. In
the northeast arm the soundings varied from 24 to 27 feet. Near the shore, except in
places where there is marsh land, the depth is from 18 to 25 feet. In the south end
of the northeast arm the land is marshy; also off the point extending into the northern
part of the lake between the two arms. In most other parts grass land extends down
to the water’s edge with a gradual slope. On the east side the land is quite level and
covered with willow and other trees. This strip of land extends back from the lake
for a considerable distance, when it suddenly runs into high bluffs.

The land surrounding the lake for the most part is high and was at one time
heavily wooded, but now only small clumps of trees are left standing, most of the once
great forest having given way to town sites and buildings. Skirting the north shore
runs the Seattle and Lake Shore Railroad. The shore is thickly settled, except on the
east and northeast sides.

The bottom in nearly all parts is mud, in some places intermingled with fine sand.

From the 2d to the 8th of December gill nets were set in ten different places,
taking 4 salmon, 1 cut-throat trout, and 2 suckers. In no place were the nets down
less than 24 hours, and in one place they remained down 72 hours. They were set in
places best adapted for this kind of fishing and where black bass had been caught
in greatest numbers. Trolling for black bass was carried on at different times, but
nothing was caught. The season was too late for them, but it was thought that the
gill nets might capture a specimen or two.

During the summer black bass are frequently taken by trolling, and it is reported
that they are sometimes taken in nets secretly set for their capture. This method is
illegal, but is carried on to some extent. Only a few black bass have been caught in
Lake Washington, and those planted in this lake in 1890 soon found their way into
Lake Union, where they are said to be quite plentiful during the summer months.
Why they should find the water in Lake Union better suited to their wants than
where they were planted is not easily explained. It can not be due wholly to the
difference in the condition of the water, for Lake Washington largely supplies Lake
Union. The theory that presents itself to account for these fish being found in Lake
Union and not in Lake Washington is that in the former lake small surface life is
much more abundant. The collections made in the two lakes show a considerable
difference in the amount taken in each trial.

In winter the temperature of water of this lake is several degrees colder than
that of Lake Washington, though ice seldom forms, and only when the winter is
unusually severe does the lake freeze over; it is probable that in summer Lake Union
is the warmer, it being shallow.

48 BULLETIN OF THE UNITED STATES FISH COMMISSION.

Temperatures taken in Lake Union.
| | Temperature.
| Date. Time. Depth. — = ==snus|} Date: Time. Depth.
Surface Bottom

Temperature.

Surface) Bottom.

1896 Feet oF. oF, 1896. Feet. OF. oF
Dee. 2 9.00 a.m ..--. 6 | 42 44 |\Dec.2....| 4.20 p.m,.-... | 24 43 42
2...) 11.00 a.m 21 42 44 : 4.45 p.m -.--- 27 42 43 |
2...| 3.20 p.m -. 12 42 42 | Heme) UirsF oie --ses 48 43 | 42 |
2 3.30 p.m 48 43 42 Bgnenl IDI pyri on Sa | 48 43 | 42
2...| 3.40p.m -. 36 43 | 42 Bocce | alcOO pam seemel 9 41 | 42
2.11; 4.00 p.m .. 24 43 42 | |
|

L a

SIUSLAW RIVER, OREGON.

In 1894 the State of Oregon erected a building on the Siuslaw River at Mapleton.
Oregon, with the intention of conducting salmon-cultural operations at that place.
This building finally reverted to the original owner of the land upon which it was
located. Believing that this station could be successfully and protitably operated, the
owner, through Mr. L. E. Bean, of Mapleton, kindly offered its free use to the United
States Fish Commission for that purpose. In response to this offer Dr. Meek and Mr.
Alexander were instructed to visit the station and make such examination as would
determine the probability of its being operated advantageously. The preliminary
inquiry, made early in September, showed that it would be necessary to eateh the fish
at the fishing-ground, several miles below the station, and retain them in boxes until
ripe. It was thought that this could be done and that the boxes could be floated on the
tide up to the station. This could be determined only through experimentation. Dr.
Meek was therefore instructed to return to Mapleton and tioroughly test the matter.

General description of Siuslaw River.—The Siuslaw River is one of the short
coastal streams in Oregon whose basin is confined entirely to the western slope of the
Coast Range. From its mouth to its source is a distance of about 100 miles. It drains
a very broken country, which is quite heavily timbered. Irom the mouth of Wildcat
Creek to near the mouth of North Fork the river flows through a canyon § to 3 miles
in width. The hills or mountains on each side of the canyon are from 200 to over 1,000
feet above the river and are very steep. At the mouth of North Fork is a large area
which is covered by water only at high tide. This region, known as the mud flats, is
on the north side of the river and includes most of the distance between Acme and
Florence. At Florence the river turns nearly north and then curves slightly toward
the west. The right-hand bank is, for the most part, a low bluff of soft sandstone;
the left is bordered by sand hills.

The Siuslaw River above the mouth of Wildcat Creek was not visited, but from
what we were able to learn that portion is quite similar in all respects to the part
immediately below that creek, except that gravel bars become much more frequent as
you ascend theriver. From the mouth of Wildcat Creek to the head of tide, a distance
of about 24 miles, the bottom of the river is sandstone in situ, over which, in many
places, are scattered large, angular, somewhat water-worn, sandstone bowlders. The
river above head of tide water is from 60 to 200 feet in width, and during the summer
in many places the water is less than 2 feet in depth.

The canyon is cut through sandstone, which in some places is nearly horizontal,
while in others it is tilted more or less, seldom more than 30°. The only portion
noticed which showed signs of being metamorphic is a bluff on the right hand of the
river between Seaton and Mapleton,

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 49

The timber in the region drained by the Siuslaw is chiefly fir, some cedar, hemlock,
spruce, maple, and alder. Some of the alders are as much as 18 inches in diameter,
and some of the fir trees a short distance west of the summit of the Coast Range are
very high. There is also in this region a very rich growth of ferns, chiefly the common
Pteris. A large portion of the timber along the lower 30 miles of the Siuslaw was
destroyed by fire some sixty years ago. This is being replaced by an undergrowth of
fir, alder, maple, etc. The timber bordering the sand hills is mostly bull pine.

The North Fork of Siuslaw River is much smaller than the main river. In the
neighborhood of Minerva the deeper holes in the river are from 20 to 50 feet wide, the
shallow places connecting these holes being about half this width. Along the river,
for a distance above tide water of about 8 miles, are many gravel bars over which the
water flows during the summer in depths varying from a few inches to about 3 feet.
The timber along the North Fork is similar to that on the main river, the under or
smaller growth immediately on the banks being some denser. Very little timber along
the North Fork below Minerva has been destroyed and searcely none above that place.
The water in North Fork is clear, though considerably stained by vegetation, much
more so than in the main fork.

The water in the Siuslaw in the summer is moderately clear, though slightly
discolored by vegetation. It also contains a fine silt or sediment of some sort, which
collects on the web of gill nets. At flood tide, about halfway between Florence and
Mapleton, the water appears of a reddish color, apparently due to the mingling of the
comparatively fresh with the brackish water, the salt in the brackish water precipi-
tating the clay sediment in the fresher water. During the summer, at the head of
navigation, the water at all times is too brackish for steamboat boiler use.

The head of tide water is about half a mile above Seaton. About half a mile
below Seaton is an island in the ri¥er, on either side of which, at low tide, the water in
the deepest place is less than 3 feet. Small steamers drawing about 34 feet of water
can not pass this island in summer except at from half to full tide.

Mapleton is 1 mile below Seaton. Below Mapleton in summer and at low tide the
water in the shallow portions of the river is seldom less than 10 or 12 feet in depth.
Seaton is properly the head of tide water, Mapleton the head of navigation.

The river below Mapleton will average at least twice as wide as it does above it,
and it varies in depth at low tide in summer from 10 to 12 feet to as much as 60 feet.

On October 16 a series of bottom temperatures from Martin Creek to Acme was
taken. The temperature of the air at 8 a. m. was 57°; surface of the water at Martin
Creek, 57°; bottom at same place at a depth of 15 feet, 58°. This bottom temperature
decreased as we approached the lower portion of the river until at Acme, in 30 feet of
water, it was 54°.

The fishermen, methods of fishing, etc.—The fishermen on the Siuslaw River are with
few exceptions Americans, Norwegians, and Indians. The fishing on the main river
is done for the most part by the Americans and Norwegians, on the North Fork
mostly by Indians. During the past season about three-fourths of the fishermen were
permanent residents on the river; the remaining one-fourth were mostly from the
Columbia River, and in general they restricted their fishing to the lower 8 miles of
the river, using drift nets. The resident fishermen are further up the river and set
their nets in the ordinary way. Each fisherman regards himself as privileged to fish

anywhere along the river he chooses. The fact that a man owns property on the river
F. C. B. 1897—4

50 BULLETIN OF THE UNITED STATES FISH COMMISSION.

is apparently no reason why he should restrict his fishery to his own water front or
prevent anyone else from fishing there. For convenience each resident sets his nets
as near his home as possible and has usually a number of well-defined sets, which are
sometimes marked. These sets are the places which he regards as the best in which
to catch fish, and are free from snags. One fisherman will seldom intentionally use
a set belonging to another.

During the past year very little attention was given to the “ Saturday night law.”
It was said that this has not only been the case in past years on the Siuslaw River,
but that it is practically a dead letter on all of the streamsin Oregon. The fishermen
on the Siuslaw seem to agree that all of them would profit as well by respecting it as
by breaking it. A few persist in setting their nets or drifting Saturday nights and,
in order that these few men may have no advantage over the others, they, too, put
out their nets on Saturday nights.

According to the law the fishing season closes on the Siuslaw River November 1.
This law, of course, is intended for all of the coastal streams in Oregon, and as the
time of the run of salmon, especially of silver salmon, varies somewhat in the differ-
ent rivers the law does not protect all alike. In the Siuslaw River the largest run of
Silver salmon occurs anywhere from October 25 to November 10. About this time the
first fall rains begin and a gradual rise in the river is accompanied by an increased run
of silver salmon. In 1896 the rains began October 24. The river did not rise much
until the night of November 7, when it rose about 8 feet, which put an end to fishing.

The largest run of silver salmon was between October 28 and November 6,
About the same condition of affairs is said to have existed last season.

The law which requires the season to close November 1 in no way protects the
chinook salmon, as the run of this fish is over by October 15. As fishing is the chief
source of income of many of the citizens on the Siuslaw River it seems to them quite
a hardship to be obliged to stop fishing during the best portion of the season. All
commercial fishing on the river is confined to that portion below tide water; in fact
very little is ever done above Mapleton until near the close of the fishing season.
The distance from the head of tide to the mouth of the river is about 28 miles.

Commercial fishing on the Siuslaw is carried on by gill nets, seines, and trolling
lines, but mostly by gill nets. The gill nets operated by the fishermen engaged in
drifting are from 100 to 150 fathoms long; they are used only on the lower 8 miles of
the river, where the current is too swift during flood and ebb tides to allow them to
be set in the ordinary way. There are a great many sunken logs and trees in this
portion of the river, as well as further up the stream, and the drifts are therefore
usually short. AI gill nets used on the river from the head of tide water to within 8
miles of its mouth are set in the ordinary way. They are all anchored by heavy rocks
tied to the lead line at distances of from 3 to 8 fathoms apart. They vary in length
from about 25 to 100 fathoms.

That portion of the river in which gill nets are set varies from about 8 to 150
fathoms in width. The gill nets also vary in length and, as no fisherman confines
his fishing to any one portion of the river, very little attention is paid to the law which
prohibits the use of gill nets reaching more than one-third distance across the river.
Nets set in the Siuslaw River reach all the way from one-fourth to the entire distance
across it. During flood and ebb tide the current in the river is so swift that nets
extending from about one-third to two-thirds across the river have their cork line
dragged so low that the net presents a comparatively small and a very poor fishing

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 51

surface. If the net extends from bank to bank the current raises the lead line
between anchor rocks and drags down the cork line to the extent of destroying a
good share of the fishing capacity of the net. It is quite evident that nets set in the
Siuslaw River fish very little except during slack water.

During the new and full moon tides the current in the river is so swift that the
outer ends of the nets, which reach only a part way across the river, are drifted toward
the shore, and they frequently become so much snagged in drifting one way that it
is impossible to pick them up until the return tide. The fishermen drifting always
take advantage of slack water, in order to fish their nets a longer time without taking
them up. In the latter part of the fishing season many nets used in drifting are
cutinto shorter lengths and set farther up the river. It seems to be a fact recognized
by the fishermen that in any one portion of the river a net will fish much better at
one bank than at the other, so it is much more advantageous to have two nets, each
reaching halfway across the river, than one single net reaching entirely across. Three
nets each reaching one-third distance across the river are often preferable to a single net
reaching the entire distance. Nets are seldom set in the river in the daytime.

Gill nets are also used in surf fishing late in the fishing season, when comparatively
few salmon are entering the river. The usual depth of the net used in surf fishing is
about half that used in drifting. At either end of the nets, which are about 150
fathoms long, long ropes are attached. The net is stretched along the shore and
allowed to be caught by the surf, which carries it some distance to sea; it is then
slowly drawn to shore after the manner of a seine. This method of fishing is so diffi-
cult and the results so small that it is not much followed.

A short distance above Acme is an excellent seining-ground, but the use of a
Seine has usually been opposed by fishermen using gill nets farther up the river. The
seine was used this season from September 10 to November 1 by those who most
strongly opposed it last year, but was less profitable than was expected. The majority
of the fishermen would favor a law restricting the fishing to the use of gillnets. The
seine was but little disturbed this year. One night a boat filled with rocks was sunk
on the ground, but this the seine brought ashore at its first haul next morning. At
another time a peculiar framework was anchored in the river, which was so constructed
as to raise the lead line when the seine was about half hauled in. This contrivance
was also brought ashore. The fact that the seine did not meet with the success
expected lessened the opposition to its use. Chinooks were caught by it for the
hatchery, and this was no doubt a factor in its favor.

Chinooks and the silver salmon are often caught by trolling in the Siuslaw River,
and they occasionally take hooks baited with salmon eggs or other bait. On October
10 Mr. L. E. Bean caught 3 chinooks and 6 silver salmon trolling for about an hour
near the mouth of Martin Creek. The following day 8 silver salmon were caught by
trolling from Mapleton to Point Terrace, a distance of about 6 miles. Some of the
chinooks and silver salmon caught in this way were red in color and the jaws distorted.
Considerable trolling is done during October between Mapleton and Point Terrace,
not only for sport, but for market as well. The amountof fishing done on the Siuslaw
River is rated by boats, each boat having ordinarily 2 meu and about 100 fathoms of
net. This past year there were from 75 to 80 boats engaged in fishing on both the
Siuslaw River and North Fork, using a total of about 10,000 fathoms of net. The
seine was operated by 6 men and 1 horse, and was equivalent to 3 or 4 boats.

liye BULLETIN OF THE UNITED STATES FISH COMMISSION.

The legal fishing season extends from September 1 to November 1. Some of the
fishermen begin fishing as soon as the law permits, but the majority do not commence
until the cannery begins operations. Tests made in August and early in September
showed that the chinook salmon did not arrive in considerable numbers until about
the 10th of September. They began to enter the river in the early part of August.
Very few were caught in the lower part of the river after the middle of October. In
the vicinity of Mapleton, however, a good many were taken as late as the last of Octo-
ber. They moved up the river apparently very slowly, evidently remaining for a
considerable time in the deeper portions of the river.

At Point Terrace the water is 56 feet deep; near the mouth of Martin Creek the
water is from 25 to 36 feet in depth. A large number of chinooks remain in these
places (or are thought to, at least) until they become red and distorted. They seem
to be very inactive. Gill nets set near these places catch very few. Fishermen say
that as soon as the first fall rains come and the water in the river freshens up a little
the fish become very active and are easily caught in large numbers. No special effort
is ever made to catch the chinooks from Point Terrace to head of tide after about
the 10th of October. Many of those taken after this date are so white-meated that

_they are worth not over 5 cents each, and they are usually thrown overboard by the
fishermen. Some fishermen think that many chinooks spawn on the margins of these
deeper holes, but there is no positive evidence of this further than that late in the
fishing season a few ripe females are caught near or in these deep holes.

October 28 and 29, just after the first important fall rains, about 50 chinooks were
caught near Mapleton, which was the largest catch at this place during the season.
A few of these were ripe and others nearly so. These fish were evidently just leaving
the deep hole near the mouth of Martin Creek about three-quarters of a mile below
Mapleton. The larger portion of this catch was rejected by the cannery.

The fall rains not only increase the amount of water in the river and freshen the
lower portion of the stream, but they also lower the temperature of the water. These
are doubtless the main factors causing the salmon to leave the deep holes and resume
their run upstream.

Salmon ordinarily reach their spawning-grounds before they are ripe, but there is
considerable evidence that they sometimes do not enter the river until almost ripe.
Mr. Leonard Christianson did considerable surf fishing in November, 1895, and
informed us that he caught a few ripe chinooks then. Eggs were forced from them
while the fish were dying on the shore. This, however, is probably unusual.

The North Fork being a much smaller stream than the main river and comparatively
few white men fishing on it, it seemed that it was of very little importance as a fishing
stream. Late in October it was learned that many chinooks were spawning all along
the stream above tide water. On October 23 and 24 the North Fork was examined
for some distance both above and below Minerva. The water was so stained that
salmon could be seen only in shallow water on the gravel bars. These when frightened
by our approach at once disappeared in deep water. A few were much mutilated and
in a dying condition; in fact, all seen were more or less mutilated. North Fork,
though not a large stream, has from tide water to a distance of about 8 miles above
a number of excellent spawning-beds, and all showed much evidence of having been
so used this year. On the 26th of October about 17 chinooks were caught, all spent
fish and all more or less mutilated. As these spawning-beds are only about 8 to 16

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 53

miles from the sea, the mutilations exhibited could not have been incurred while on
their way from the sea, but were received on the spawning-beds. Several dead ones
were seen, and many others were so weak and injured that recovery was not believed
possible.

On October 23 and 24 at least 50 chinooks were seen on spawning-beds. On all
of these the fins were considerably frayed, while most of them were badly mutilated
and covered with fungus. A few were so far gone that we captured them by wading
in the water and catching them by the tails.

The seals collect in large numbers during the salmon run on the bar at the mouth
of Siuslaw River. Some enter the river and are seen nearly half way to Mapleton.
They are regarded as being destructive to the fishing interests, both by destroying
large numbers of fishes and preventing many more from entering the river. These
seals are regarded by the fishermen as by far the worst enemy of the salmon.

The run of silver salmon began later. The first examples were taken about Sep-
tember 10, but they did not appear in any considerable numbers until about a week
later. From that date they continued pretty evenly until the 20th of October, after
which they appeared in muck larger numbers. They were still abundant on November
7, when a sudden rise in the river washed away many of the nets and practically put
a stop to all fishing for the season.

The importance of tle salmon fishery of the Siuslaw River for each year since its
beginning in 1889 is shown in the table which follows. For convenience of reference
the salmon taken in Tsiltcoos (Ten-mile) Creek are included.

Siuslaw River. | North Fork. Tsiltcoos|
Year. = = S| OR
Chinook,| Silver. | Chinook.| Silver. silver.
eee BA ot ES = Sas EES eS SATUS}
9,000 | 62,340 4,000 16, 432 | 8, 000
6,500 | 41,320 | 2,340 9, 320 (a)
16,500 | 23,450 6, 800 2,100 | (a)
7,000 | 86,340 38, 050 13, 460 (a)
4,000 | 78,430 | 1, 100 12,000 | 6,300
| 1,300 | 84, 642 1, 234 6, 742 4, 870
| 4,200 | 69,454 | 1,600 4, 340 8, 200
| 4,378 57, 000 2, 100 4, 650 1,578
|
| Motalstose seen ee eee 52, 878 | 502, 973 22, 224 69, 044 27,948

a Did not fish.

From this it appears that the total number of chinook salmon from the Siuslaw
River and the North Fork handled by the canneries in the eight years was 75,102 fish,
or 1,301,632 pounds. Mr. Kyle estimates that this amount should be increased by
one-sixth, to cover the number salted or shipped away. This would give 87,619 fish,
or 1,518,750 pounds. The total number of silver salmon handled during the same
period was 599,965 fish, or 5,429,675 pounds; increased by one-sixth for fish salted or
otherwise utilized, the total becomes 699,959 fish, or 6,334,620 pounds. The number
of silver salmon taken in the outlet of Tahkenitch Lake (Five-mile Creek) is said to
have averaged about 5,000 fish per year for the last three years.

Condition of salnon.—The chinooks caught in the Siuslaw River are not so fat as
those caught in the Columbia. This is no doubt due to the fact that the Siuslaw has
only a fall run. In any river the spring run is always better than the fall run; the

54 BULLETIN OF THE UNITED STATES FISH COMMISSION.

early spring run is better than the late spring run, and the early fall better than the
late fall run. The condition of the salmon depends on the length of time he remains
in fresh water before spawning. Those whose sexual organs are the least developed
have redder and fatter flesh, and these are the ones which must remain in fresh water
the longest time before spawning.

Salmon taking the hook.—Fishermen hold different opinions as to why salmon take
the hook, some regarding it simply as due to the pugnacious disposition of the fish,
others as showing a playful disposition. Both males and females will take the hook.
The baited hook falling near them or the silvery spoon passing rapidly by no doubt
produces an involuntary impulse on the part of the fish to seize it.

The Siuslaw hatchery.—This hatchery is at Mapleton, Oregon, on the Siuslaw
River, about 26 miles above its mouth, or 14 miles below the head of tide water. The
building is about 300 feet from the river and is supplied with water by gravity from
a smnall spring brook, which flows within 200 feet of the building. This supply is
abundant and of good quality. As there was no rack in the river, fish for spawning
purposes could be obtained only at the seining-ground near Acme, about 16 miles
below Mapleton. The method followed was to retain the fish in live-boxes until they
were ready to spawn. Special boxes were built for this purpose, the fish put in them
at the seining-ground, then they were floated on the tide up to Mapleton, where they
were tied until the fish were utilized.

The first live-box was built September 17. It was made 20 feet long, 10 feet wide,
and 44 feet deep. The frame-work was made of 3 by 6 material; to this were nailed
horizontally 1 by 4 strips, with interspaces of about 24 inches. The lumber used was
undressed. This box will hereafter be referred to as Box No.1. On September 19
Box No. 2 was built, 16 by 8 by 44 feet. This box differed from No.1 only in being
made out of boards 6 and 8 inches wide instead of 4 inches, and in having the inter-
spaces narrower. These were not more than 2 inches wide anywhere, and near the
top of the sides no interspaces were left, the first two or three being placed against
each other. On September 21 Box No.3 was built, not differing in any way from Box
No. 2. On the 1st of October Box No. 4 was built. It was made of dressed lumber
and was 20 by 64 by 43 feet. Care was taken to have this box as smooth inside as pos-
sible, so that the fish might not become injured in any way on account of contact with
the box.

The account of the experiments with each of these boxes is given with considerable
detail, in order that the nature of the work and the conditions under which it was
done may be fully understood. On September 18,43 chinook salmon were placed in
Box No. 1. They were obtained from the seines and carried in sacks a few feet to a
small live-box, in which they were floated to No.1. The next day 48 more salmon were
placed in this box. On September 21, 47 salmon were placed in No. 2, and the next
day 22 fish were put into No. 3, to which 6 more were added on the morning of the
23d. These three boxes, with 91, 47, and 28 fish, respectively, were lashed together
and were started drifting up the river with the tide. This tide carried them upstream
about 7? miles, or about 4 mile below McLeod’s wharf. Here the boxes were taken in
tow by the steamer Lillian and brought to McLeod’s wharf. The next day they were
drifted to Point Terrace, a distance of about 54 miles, and on the following night they
were drifted 25 miles farther to Hartley’s wharf. The next day, the 25th, they were
drifted 13 miles farther, which brought them within } mile of Mapleton. They were

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 55

tied up here until the 27th, when the Lillian took them in tow and brought them to
the Mapleton wharf, where they were kept until the fish were spawned.

Three fish were placed in Box No. 4 on the evening of October 1, and 29 more were
added the next morning. This box was set drifting on the evening of October 2.
The fish were quite restless, jumping and striking against the top of the box in efforts
to get out. After floating about 14 miles this box was tied up for the night. That
evening it was covered with spruce boughs, so as to darken it and at the same time to
cause it to float lower in the water. The next morning it was drifted to within half a
mile of McLeod’s wharf, or about 6 miles. The box was anchored here until the even-
ing flood tide, when it was drifted to McLeod’s wharf and tied up. The next day it
was drifted about 65 miles to Hartley’s wharf, and on the day following it was floated
to within half a mile of Mapleton, to which place it was taken on the next tide on the
morning of October 6. It was tied in a protected place in the river opposite Mapleton,
where it was not subject to the disturbances to which the other three were liable. On
October 8 two more salmon, which had been caught in a gill net near by, were placed
in this box, thus increasing the number to 34. Except during the first evening the
fish in this box remained quiet and showed no signs of restlessness under the restraint.
Boxes 1, 2, and 3 were tied to the piling at the upper end of the wharf at Mapleton.
The fish were here occasionally disturbed by persons coming about the boxes or step-
ping upon them; whether this really increased the mortality or not can not be certainly
stated, but it seems reasonable to believe that it would prove detrimental to the fish.

An unusually low tide on October 15 left Box No. 3 about one-third out of water.
This seemed to distress the fish, and may have caused some injury. During the latter
part of October about 12 more fish were put in No.1. These had beer caught in the
gill nets at Mapleton.

The total number of fish which were experimented with was as follows: In Box
No.1, 103; in No. 2, 47; in No. 3, 28; in No. 4, 34. No accurate record was kept of
the number of each sex, but at least three-fifths of the total number were females.

The mortality among the fish in these different boxes is shown in the following
tabular statement:

Date. Box No.1. | Box No. 2. | Box No. 3. | Box No. 4.
|

DAO 7 etna hes ee 8

Total loss. ---.-.......... 60
Percentage of loss.......-...-. 58

After the taking of spawn began, October 26, the fish were shifted about from one
box to another, and, though a few continued to die, no accurate record was kept. The
mortality was greatest in No. 1 and least in No. 4. The crowded condition of the fish
in No. 1 was doubtless a feature which contributed to the loss.

After October 26, when spawn-taking began, the fish were shifted about from box
to box, and it was therefore impossible to keep an accurate record of the number dying

56 BULLETIN OF THE UNITED STATES FISH COMMISSION.

in each as originally apportioned. The total number that died after October 27, how-
ever, was 18, and on November 15 a sudden rise in the river broke open box No, 1
and 30 fish escaped. The total loss, therefore, was 112 dead and 30 escaped. The 70
which were left were spawned. Of this number, 36 were females and 34 males; to
this number should be added 8 females and 8 males taken from the gill netsat Mapleton;
these 86 (44 females and 42 males) were spawned at different times between October
26 and November 15. The total number of eggs taken is stated by Mr. L. E. Bean to
have been 217,000.

An examination of a great many of the fish which died showed that some had
received internal injuries which probably caused their death. Others which showed
no internal injuries exhibited severe bruises which doubtless hastened their death.

Particular pains were taken with the fish in Box No. 4. Care was taken in
handling them when they were first placed in the box. The box had been constructed
with unusual care by making it smooth inside and placing the boards close together,
so as to admit but little light, and it was anchored at Mapleton in a secluded place
where the fish were subjected to the minimum amount of disturbance. <As a result, a
higher percentage of success was attained with this box than with the others. With
Nos. 1, 2, and 3 the principal factors which were detrimental were the following:

1. Too many fish were put in each, resulting in too close crowding.

2. The interspaces were too wide, thus letting in too much light and making the
fish more restless.

3. The rough surfaces on the inside of these boxes caused injuries to the fish
striking against them.

4, Some of the fish were not handled with proper care when placed in the boxes.

5. During the interval between the arrival of these boxes at Mapleton and the
spawning of the fish they were kept in a place where the fish were subjected to some
annoyance.

If the fish in these three boxes had been treated with the same care and attention
as was bestowed upon Box No. 4, it is quite certain that the measure of success would
have been greatly increased. Of the 217,000 eggs taken Mr. Bean reports that he
hatched and successfully liberated 180,000.

Conclusions.—These experiments seem to demonstrate that it is entirely practi-
cable to retain salmon intended for spawning purposes for a period of at least six
weeks in properly constructed live-boxes, but in order to operate Siuslaw hatchery
with the best results, particular attention should be given to the following details:

1. The boxes should be constructed with as much care as possible. They should
be made smooth inside. No interspaces should be left in the top nor in the upper third
or half of the sides and ends, and those elsewhere should not exceed 2 or 3 inches in
width. It is also desirable that the boxes be made 6 to 8 feet deep instead of 44 feet.

2. It will prove advantageous to get as large a proportion as possible of the fish
desired toward the close of the run, as they will be more nearly ripe and will not have
to be kept so long in the boxes. It is also believed that these fish are not so restless
in the boxes as are those caught earlier in the run.

3. The fish should be handled carefully when placing them in the boxes, so that
they may not receive any serious bruises,

4, Too many fish should not be placed in any one box. In our judgment each fish
should be allowed at least 20 cubic feet of space.

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 57

TSILTCOOS AND WHOAHINK LAKES, OREGON.

The investigation of these lakes was undertaken in response to a petition signed
by numerous citizens living in their immediate vicinity and addressed to the United
States Commission of Fish and Fisheries under date of June 13, 1896. The petitioners
asked that Tsiltcoos Lake be stocked with black bass and that a plant of brook trout
be made in Whoahink Lake. To determine the advisability of complying with these
requests, Dr. Meek was directed to make an examination of these lakes, particularly
in regard to the following points:

1. Are these waters suitable as to temperature, quality, and food supply for the
species requested ?

2. Would the introduction of these species have any detrimental effect upon the
food or game fishes native to those waters or to waters into which the introduced
species might spread ?

3. Would the species which it is proposed to introduce probably thrive so well as
to afford, after a few years, better fishing than is now furnished by the native species?

Three visits were made by Dr. Meek to these lakes, the first on October 7 and 8,
when only Tsilteoos Lake was seen; the next October 15 to 19, during which time
considerable work was done on each of the lakes; the third trip November 28 to
December 7, when the lakes were examined carefully and extensive collections made.

The conclusion reached from a study of the conditions obtaining at these lakes is
that it is not advisable for the Commission to stock them with black bass or brook
trout. The planting of black bass in Tsilteoos Lake would prove detrimental to the
silver-salmon fishery carried on in the outlet of that lake, and the probability that
the bass would in time spread to the Siuslaw and the Umpqua is too great to warrant
the risk. As to placing brook trout in Whoahink Lake, it is not believed that that
species would ever become so well established as to afford better fishing than the
native species supplies. These lakes are already abundantly supplied with a native
trout which attains a weight of 2 pounds or more, which possesses excellent game
qualities, and whose flesh is firm and sweet. If these trout are properly protected,
there is no reason why Whoahink Lake should not become an important fishing-resort.

The detailed report upon these lakes follows:

TSILTCOOS AND OTHER LAKES.

Between the mouths of the Siuslaw and Umpqua rivers, and almost bordering
the sand hills along the shore, are three large lakes and a few small ones. The invyes-
tigations were restricted to the large lakes; these lie in a line parallel with the ocean
beach and comprise at least three-fourths of the distance between the Siuslaw and
Umpqua rivers. They are quite peculiar on account of their great irregularity in
outline. The amount of shore line as compared with the area of each is very great.

Whoahink Lake, or Clear Lake, as it is more commonly known, is about 2 miles
from the Siuslaw River at Glenada, This lake is very irregular in outline and deeper
than the other two. Except a small portion in the southwest corner, it is surrounded
by hills from 50 to about 200 feet in height. The hills immediately surrounding the
lake are composed of a soft, irregularly stratified sandstone, and evidently of a much
newer formation than the higher hills to the southeast. The inlets are only ravines,
some of which contain many small springs. The points of land projecting into the

58 BULLETIN OF THE UNITED STATES FISH COMMISSION.

lake are being quite rapidly worn away by rough water, forming blutfs. The surface
of the lake at one time was somewhat lower than at present; this is evident from the
fact that on a shoal place on the east side (near the middle of section 14) stumps of
trees are seen some 10 feet below the surface. The former outlet of the lake was into
the outlet of Tsilteoos Lake. A few years ago this outlet became closed by shifting
sand, causing the water to rise in the lake at least 6 feet above its present level. A
new outlet, a short distance to the east of the old one, was dug to Tsiltcoos Lake
some four or five years ago. The new outlet was not only intended to lower the sur-
face of the lake, but to afford a water power for a smallsawmill. It has never been so
used. The material through which the outlet was dug is a fine clay, so compact that
it erodes very slowly. In the extreme ends of the arms of the lake are some tules
and small areas of other water-plants.

The surface of Whoahink Lake is about 15 to 20 feet above that of Tsiltcoos Lake,
from which it is less than half a mile distant. The timber on the hills bordering the
lake was nearly all destroyed by fire some sixty years ago. It is being replaced by a
growth of fir, alder, hemlock, maple, rhododendron, huckleberry, and a vigorous growth
of ferns. To the west and bordering the sand hills about half a mile distant from the
lake are many bull pines. The shore was everywhere so snaggy that a seine could not
be used and our collecting apparatus consisted only of gill nets and trot lines. This
lake is reported as having very few fishes in it, and our experience confirms this view.
The water is clear, though much less so than in the mountain lakes of Idaho; the
depth is usually from 30 to 78 feet. The water is not very cold, as may be seen from
the following recorded temperatures:

—— =

Tem- | Temperature of water.
Date. Hour. DOr lite 3
| ature Sur- | Bot-
| of air. face. | Depth. tom.
1206. _| oF, | oF, | Feet. | OF. |
Oct. 18 | 4.30 p.m. 63 625 |e 61
AB oe vest ncot Paal a etm, [Eee oe 72 | 56
19 9.30a.m.| 58 | 61.5 30 | 60
19) bee onaceesoceces|cesccke nee oesee 54 | 61.5
19)|:2o--seneeu sees | Pets an ee aa 66 | 57

Tsiltcoos lake is larger and more irregular than Whoahink Lake. The main body
of the lake is about 2 miles in extent from north to south, and 1 to 2 miles from east
to west. It is also supplied with several arms, usually about a fourth of a mile wide
and from 4 to 14 miles long and extending in different directions. Tsiltcoos Lake is
comparatively shallow. In summer its greatest depth is about 17 feet, its average
depth being from 10 to 14 feet. During the rainy season the water is from 6 to 10
feet deeper. The bottom is a soft, black mud, like that of Whoahink and Tahkenitch
lakes. Except at a few intervals the shore is bordered by a rich growth of tules. In
the shallow portions around the shore the tules extend into the lake about a fourth of
amile. There is also considerable swamp vegetation in and about the tules. The
surrounding country is quite similar to that around Whoahink. On the east, and at
one point on the south, are bluffs of a dark, compact sandstone of much older formation
than the bluffs around Whoahink Lake. There are also two timbered islands in this
lake, composed of this older sandstone; the larger is about half a mile long and about
an eighth of a mile wide. That portion of the country bordering the lake on the south
and southwest was not burnt over during the big fire of about sixty years ago,

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 59

Tsiltcoos Lake has two inlets, 6 to 8 miles long, Maple Creek at the northeast and
Tsiltcoos Creek to the southeast corner of the lake. In summer these are little more
than small brooks, but during the rainy season they become much swollen and afford
a spawning-ground for many silver salmon, and possibly a few others, which enter the
lake. It is said that about 6 or 7 years ago many more salmon spawned in Maple
Creek than at present.

The outlet of Tsiltcoos Lake is quite small, especially so in summer and at low tide.
Gill nets are put across the outlet at low tide and are so anchored and staked down
that it is quite impossible for silver salmon, except the smaller ones, to pass them.

October 17 a number of soundings and temperatures were taken in Tsiltcoos
Lake. The temperature of the air at 10.15 a.m. was 56°; that of the bottom in various
parts of the lake, at depths from 7 to 17 feet, varied from 58.5° to 59.59,

Tsiltcoos Creek (Ten-mile Creek), the outlet of Tsiltcoos Lake, is about 5 miles
long, while the distance from the lake to the ocean in a direct line is about 2 miles. At
low tide in summer there is barely water enough in the outlet to float a small flat-
bottom boat. At high tide the water in the outlet for its entire length is brackish.
There is some commercial fishing, though the fishing season is short (this year from
October 10 to about November 5). Many silver salmon enter the lake, as is evidenced
by the number caught in our small gill nets. Fishermen report that salmon are often
stranded on the beach. About 100 silver salmon, the first catch of the present season,
were brought to the cannery on October 15. There is no reliable evidence that blue-
back salmon are ever found in this lake or its outlet, and only an occasional chinook
js found. On October 16 and 17 the bar at the mouth of this creek at low tide com-
pletely closed the outlet. This prevented salmon from entering and none had been
taken since October 13. The fishermen were watching the surf closely to see if many
salmon were near shore. Ordinarily, if the bar is closed and many fishes are seen in
the surf, the fishermen resort to surf fishing until after the bar is opened.

Tahkenitch, or Five-mile, Lake is very irregular in outline. Its greatest length
from north to south is about 44 miles; east to west about the same distance. This
lake consists of four large arms whose general directions are with the four cardinal
points. The eastern arm is the longest, the western the shortest, while the other two
are about equal in length. The width of these arms varies from about one-eighth
to three-quarters of a mile. In the mouth of the smallest arm is a small timbered
island.

Tahkenitch Lake does not lie in the burnt region. The hills surrounding it rise
abruptly to a height of usually about 100 feet, and are covered by a dense growth of
tall fir trees of about 14 to 3 feet in diameter. In the extreme end of the arms and
in a few other places along the shore are found tules and other water vegetation similar
to that in Tsilteoos Lake. The bottom of the lake is a soft, black mud, like that found
in the two other lakes. The depth is very uniform, being from 16 to 22 feet.

Five-mile Creek, the outlet of Tahkenitch Lake, has its origin in the distal end of
the western arm, its upper end being about 1 mile from the ocean. Its total length is
about twice this distance. The sand hills extend almost to the lake. The isthmus
separating Tahkenitch and Tsiltcoos lakes is rather low and narrow. So far as we
could learn, the commercial fishing on Five-mile Creek is very similar to that on
Ten-mile Creek. The catch has amounted, according to Hon. A. W. Reed, of Gardiner,
to an average of 5,000 silver salmon for each of the past three years.

60 BULLETIN OF THE UNITED STATES FISH COMMISSION.

UPPER KLAMATH LAKE, OREGON.

In 1889 the United States Fish Commission planted 400,000 whitefish fry (Core-
gonus clupeiformis) in the southern end of Upper Klamath Lake, near Klamath Falls.
Ample time having elapsed to enable this species to become well established, observa-
tions were made by Messrs. Meek and Alexander for the purpose of determining the
result of the plant.

Pelican Bay was reached on the evening of October 31, and work was carried on
in the upper part of the lake until November 7, when the camp was moved to the lower
end of the lake near the outlet, where the investigations were continued until the 11th.
During the time spent at this lake the weather was unusually stormy, and it was
exceedingly difficult to carry on the investigations in a manner at all satisfactory.
There were only two really good days for work during the entire time.

For the purpose of determining the result of the whitefish plant gill nets of suitable
mesh were set in various places in Pelican Bay and in the lower end of the lake.
Collections of native fishes were also made, and as much time as possible was given to
the study of the fish-food supply of the lake, which was found to be abundant and rich
in species of Hntomostraca, insect larve, and other invertebrate life.

The fish life of the lake does not consist of many species, but the individuals of
several of the species are very numerous. These will be discussed fully in the list of
fishes at the end of this report, but it seems proper to call attention in this connection
to the suckers and trout. There are five or six species of suckers in this lake and
each of them attains a large size, which renders them of great importance to the
Indians on the Klamath Reservation. In the spring of the ) car, during the spawning
time of these fish, vast numbers are caught in traps and by hook and line by the
Indians and cured for future use.

The trout are also very common and reach a very large size. Trout weighing 8
to 10 pounds are not at all unusual, examples of 10 to 12 pounds are not rare, and
occasionally one weighs as much as 14 pounds. Trout can be taken readily with the
fly in the spring and by trolling in the spring and fall. In Pelican Bay they may be
taken by trolling at any time.

The Klamath lakes comprise a series of lakes which extend in a north-and-south
direction in southern Oregon and northern California. They lie just east of the
Cascade Mountains, and are about 4,300 feet above sea level. Our investigations
were limited to the largest of these lakes, known as Upper Klamath Lake.

This lake is about 25 miles long and about 8 miles in average width. Its northern
half extends in a northwest-and-southeast direction, while its southern half is nearly
due north and south. It is quite irregular in outline, having on its western side two
large bays, the upper of which forms the northwestern corner of the lake, and is
known as Pelican Bay. About 2 miles west of Pelican Bay is a very large spring
which forms a prominent creek emptying into the bay. The first half mile of this
creek is about 50 to 100 feet wide and from 2 to 8 feet deep. After this it widens out
into a large estuary, half a mile or more wide and more than a mile long. On the north
this estuary is bordered by a hill, at the foot of which the water is as much as 17 feet
deep. The rest of this estuary is bordered by a rich growth of tules and swamp
grass. Its depth, except at the north end, seldom exceeds 7 feet, while the greater
portion is less than 6 feet in depth. There are, especially on the east side of this
estuary, many small narrow arms known as sloughs. These sloughs are about as

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 61

deep as the estuary, and each supports a rich growth of Potamogeton, Myriophyllin,
and Hlodea. These water-plants are also more or less abundant along the water’s
edge just inside the tules, but are scarce in the main body of the estuary. The water
in Pelican Creek and estuary is very clear, and when it is not disturbed the bottom
can be plainly seen everywhere. The bottom of the creek and estuary is composed
of a light fluffy mud, varying in color from dark grayish to black.

The water in the Jake when we saw it was not clear. The fine sediment held in
suspension was no doubt picked up from the bottom during rough weather.

The northern portion of Upper Klamath Lake contains many tule islands, and is
evidently very shallow. Its shores north of Bare Island are very marshy. The main
portion of the lake, or its middle half, is bordered on each side by mountains which
rise from the water’s edge to an altitude of 1,000 to 1,500 feet above the lake. Toward
its southern end the lake becomes narrow. Its shores are mostly marshy and bordered
by tules and willows. It also contains considerable water vegetation, the same as at
its north end. There are two islands in the lake, Bare Island and Buck Island, each
about 500 feet high.

North of the upper end of Upper Klamath Lake is a large area of tule marshes,
6 to 10 miles wide and 10 to 12 miles long. Lying in this marsh is a small lake which
has sometimes been improperly called the Upper Klamath Lake. It is now regarded
as being only a part of Upper Klamath Lake. It is about 2 miles from the main lake
and is about 3 miles wide and 6 miles long. Through this marsh flow several small
streams, the principal ones being Seven-mile Creek and Wood River. Williamson
River, the principal stream flowing into Upper Klamath Lake, is of considerable size
and much importance in its relations to the fishes of the lake. About 10 miles above
its mouth it receives from the east a large tributary known as Sprague River. About
half a mile above its mouth if is 125 feet wide and 10 feet deep. It flows before
entering the lake through a low, flat, marshy plain covered with tules and swamp grass.
On the banks of the river is a rich growth of willows. The current is very moderate.
We rowed about 2 miles or more up the river and noticed no important change in it.
The water in the river was very dark, evidently stained to some extent by vegetation.

The small lake just below Klamath Falls is about # mile wide and 2 miles long.
This lake is surrounded by a rich growth of tules and some willows. The depth about
half a mile below Klamath Falls was 15 feet. On November 9 the surface temperature
was 424°; bottom temperature in 15 feet of water, 424°. Weset one large and two
small gill nets in this lake on November 9 and took them up on November 11, obtaining
4 chubs (Leueiscus bicolor) and 2 trout.

The outlet of Upper Klamath Lake is Link River, a short stream not over 14
miles long, consisting of a series of rapids and falls.

Upper Klamath Lake is a rather shallow body of water. The creek and estuary
at Pelican Bay are from 2 to § feet deep in the upper part and from 6 to 17 feet in the
lower portion. A line of soundings was run across the mouth of the estuary and the
depth was found to be about 54 feet.

Pelican Bay and the north end of the lake are very shallow. A line of soundings
was run from the mouth of the estuary of Pelican Creek to the mouth of Williamson
River, and 8 feet was the greatest depth found, while the average was not more than
63 feet. The bottom could be easily seen\anywhere when the surface of the water
was smooth. It seemed to be composed of loose, decaying vegetation, into which the

62 BULLETIN OF THE UNITED STATES FISH COMMISSION.

sounding lead would usually sink several inches. We were unable to take any sound-
ingsin the middle of the lake between Bare and Buck islands. While this is doubtless
the deepest portion, we were informed that the water even here probably does not
exceed 30 or 40 feet in depth. The south end of the lake is a little deeper than the
upper portion. The greatest depth was found near the outlet and was 20 feet. The
portion south of Buck Island will not average more than 12 or 15 feet deep.

The temperature observations made at Upper Klamath Lake are given in the table
which follows. All the water temperatures except those at the surface were taken at
the bottom. Owing to the shallowness of the lake the temperature of its water is
everywhere affected by that of the air.

Tem- ais | Tem-
F . per- ept per-
Date. Hour. Station. ature of of water. ature of
air. water.
Pint: oF,
Nov. 2 -| Near mouth of Pelican Creek. - 38 | Surface. | 42
2 North end of Pelican Estuary |.---do...| 40
2 Pelican Bz 53 feet .. 414
2 -do .- 54 feet .. 393
OV Mfehol i) Aire es Ie Gee doses 5g feet..| 393
3 Pelican Creek near sp Surface - 44
3 -.| Pelican Estuary -..------ cai =5- 41
3 - Pelican Creek near spring do .- 44
3 Pelican Estuary - do . 42
4 Pelican Creek near spring do -. 44
5 Pelican Creshke soves se eee eee eee eae aoe eee do.. 433
5 | Northwest part of Pelican Estuary - 2 SeGOres 403
OY PO aa ee ee ere oe CUE See ae a ene Ba ees ee ee sce “| 16 feet .. 41}
5 Klamath Lake between Pelican Bay and Wil- |-.--..--- Surface -| 424
| | liamson River. |
5 | N 7 feet -.. 43
5 | 12.3 --| Surface . 42.
5 | 22 --| 8 feet ... 43
5 | --| Surface - 422
5 --| 8 feet ... 42
5 --| Surface - 42
5 --| 10 feet -. 42
5 --| Surface - 423
5 | --| 10 feet .. 42
6 -| Surface - 394
6 5yfeet..| 404
9 Surface - 42
9} 15 feet -. 42
10 --| Surface - 413
10 -| 13 feet 2 413

Character of bottom.—The large area of tule and marshy lands surrounding the
lake has doubtless had much to do in determining the character of the bottom of Upper
Klamath Lake. The lake is very shallow, and the vast amounts of decaying vegeta-
tion carried into it and the ever-increasing area of tule lands render it more shallow
year by year. The bottom, where we examined it, was composed of loose, disintegrat-
ing vegetable material, with no sandy or gravelly bottom.

Fishes.—Gill nets of suitable mesh were set in this lake in various places, chiefly in
Pelican Bay and in the lower end of the lake. A good many trout and chubs and a
few suckers were taken, but no whitefish were caught and none was seen anywhere.
Inquiry among people at Klamath Falls and elsewhere about the lake failed to elicit
evidence that whitefish have ever been seen there since the plant was made. It is
to be regretted that nets could not have been ‘set in the middle portion of the lake,
as it is there that fish would be found if any have survived. But when it is consid-
ered how shallow the lake is, how warm the water probably becomes in summer, and
how difficult it would be for whitefish to find suitable spawning-beds, it is doubtful if
the eastern whitefish would thrive in such a lake.

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 63

Advantages of a trout-cultural station at Klamath Falls —The trout of Upper
Klamath Lake are probably not surpassed by any lake-dwelling trout in America in
beauty, size, gaminess, or sweetness and delicacy of flavor They are excellent trout
to propagate and plant in other lakes. Trout-cultural operations could be carried on
at some point on Upper Klamath Lake very advantageously. DPerhaps the best loca-
tion would be at Klamath Falls, where all the engineering conditions are excellent.
Fish can be obtained readily and in abundance at any time. <A station located upon
this lake would be useful in maintaining the supply of trout in it and neighboring
lakes and for furnishing fish for planting elsewhere in suitable lakes. The distance of
Klamath Falls from the railroad is the chief objection to the establishment of a station
at that place.

CRATER LAKE, OREGON.

In the spring of 1896 the U. 8S. Commissioner of Fish and Fisheries received a
request to plant trout in Crater Lake, Oregon. This request was made by citizens
of Medtord, Ashland, and Klamath Falls, and by the Mazamas, a club of mountain
climbers with headquarters at Portland, whose members take an active and intel-
ligent interest in discovering and making known the scenic and other natural
attractions of the Northwestern States. Before complying with the request, the
Commissioner decided to haye made such investigations as would determine whether
the physical and biologic conditions existing at Crater Lake are such as will permit
fish to thrive init. These investigations were made by Professor Evermann and Prof.
U. O. Cox, of Mankato, Minnesota.

Crater Lake is about 100 miles from the nearest railroad station, and is reached
by a wagon road, which has a number of steep, rough grades. The Mazamas had
selected this lake as the place of their annual meeting for 1896, and the Fish Com-
mission party by their invitation made the trip to the lake with them, thus obviating
the necessity of employing special conveyances and making the trip comparatively
inexpensive. The party left Ashland by wagon August 13 and reached the lake on
the morning of August 19. Work was carried on at the lake until August 24, when
the party set out upon the return trip to Ashland, which was reached August 26.

Crater Lake* lies in the top of Mount Mazama and on the very summit of the
Cascade Range, about midway between Mount Shasta and Mount Hood. Its latitude
is 42° 56’ N. and its longitude is 122° 7’ W. In many respects it is one of the most
interesting natural wonders in America. It is approximately circular and averages a
little more than 5 miles in diameter. It is completely encircled by a bold escarpment
ranging from 500 to 2,000 feet in height above the surface of the water. Although
the steep slopes of the escarpment are in some places well wooded, they are generally
either cliffs or talus, descending to the lake and plunging into deep water. There are
many places where the walls are almost perpendicular, and at only two or three places
is it possible to descend to the water’s edge. The lake has practically no shores or
beaches; only in a few places is there sufficient beach to afford standing room. The
average diameter of this great pit at the top is 5.7 miles and its depth is 4,000 feet.
The highest part of the wall surrounding this lake is 8,228 feet above sea level, while
the surface of the lake is 6,239 feet above the sea.

*In our general description of this lake we have made free use of the excellent account given
by Mr. J. S. Diller, of the U. S. Geological Survey. Mr. Diller has made a careful topographic survey
of the region and possesses a thorough and intimate acquaintance with Crater Lake.

64 BULLETIN OF THE UNITED STATES FISH COMMISSION.

The rim of the basin now occupied by Crater Lake is the present summit of what
was once a mountain cone. That this is simply the frustum of what was once a com-
plete mountain cone is evidenced by the attitude of the sheets of lava and ejected
voleanie material which form the rim. They all incline away from the lake, indicat-
ing a common source from a crater that surmounted a huge volcano high in air above
the place now occupied by the lake. At that time, of course, neither basin nor lake
existed. Mr. Diller thinks that this great voleanic mountain, recently named Mount
Mazama, must have extended about 5,000 feet higher than the remaining fragment,
thus making its height about 14,000 feet, nearly that of Mount Shasta. What became
of the top of Mount Mazama is an interesting question. The geologists are convinced
that at some period in its history the molten interior, instead of escaping through the
crater in the usual way, found an outlet at a lower level. The top of the mountain
being thus left amere shell and without adequate support, fell in and became engulfed
in the vast pit which the lake now partly fills. The pumice upon the surface for
many miles around Crater Lake was probably blown out by the volcano before the pit
developed, and the voleano of Wizard Island was active at a much later stage upon
the bottom of the pit. It was the scene of the last eruption about the lake, and,
although recent in appearance, must have occurred centuries ago.

Crater Lake is, so far as known, the deepest lake in America. Many soundings
have been taken by the United States Geological Survey. Over oniy very limited
areas was a depth of less than 1,000 feet found, while over a large part of the lake the
depth ranges froin 1,800 to 2,000 feet. In the eastern portion of the lake is a great level
floor, with a nearly uniform depth of 2,000 feet for a distance of more than 3 miles.
In the western part are three or more small cones, one reaching within 93 feet of the
surface of the water, while another rises 845 feet above the water. This is Wizard
Island, a perfect cinder cone, with symmetrical slopes, and in its top a crater 80 feet
deep. This cone is composed chietly of red lapilli, and so new and fresh that it is
sparsely forested and shows no trace of weathering. About the base of this island
cone is a rough fringe of lava, which has spread in all directions to far beneath the
water’s level. The shallowest parts of the lake lie west and south of Wizard Island.
The average depth in this area probably does not exceed 150 feet.

Crater Lake has neither inlet nor visible outlet. The catchment area is scarcely
greater than the lake itself. During the summer season several streams trickle down
the walls from the snow banks which lie within the rim; these streams are all very
small, but a few of them continue throughout the summer, as some of the suow banks
lying on the southern walls never entirely disappear. The amount of precipitation is
believed to be greater than the evaporation, but the level of the lake does not appear to
be rising. It is therefore quite probable that there is a subterranean outlet, and the
large springs in the vicinity of Fort Klamath may have their source in Crater Lake.

The water of Crater Lake is cold, fairly pure, and exceedingly clear. Though
sufficiently pure for all fish-cultural purposes, it is slightly flat to the taste and doubt-
less contains some minerals in solution. Froth or foam readily forms when the water
is disturbed. The clearness is remarkable, an ordinary dinner plate being distinctly
visible at a depth of nearly 100 feet, even in a hazy atmosphere.

The color of the lake is a wonderful blue, except in certain comparatively shallow
places near the shore. In the deeper parts of the lake it is the most intense blue we
have ever seen; from there toward the shore, and in the changing light and shadows of

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 65

cliffs and clouds, the colors change from ultramarine through cobalt and azure blue to
smalt blue and hyacinth, and even to royal purple, violet, and mauve. So marvelously
and strangely beautiful are these colors that one never tires watching and studying
them. The general effect as one views the lake from some advantageous point upon
the rim is profoundly impressive. Two thousand feet below lies the lake in whose
placid blue waters everything is so perfectly mirrored that one can not tell where the
realends and the mirror begins. Near the west shore rises Wizard Island, symmetrical
and beautiful in all its proportions, while around the lake is a circle of 20 miles of
nearly perpendicular wall, hundreds of feet high and unrivaled in its scenic effects.

Temperatures.—Considerable time was devoted to the making of temperature
observations in different parts of the lake and at different depths.

The following table gives the surface temperatures recorded :

= |
Temper-

Date. Hour. Station. ees |
1896. | | bi
Aug. 19 From shore in Eagle Cove.....--..-..-.--- 59
20 One-fourth mile from shore in Eagle Cove. 57 |
20 | One mile from shore in Eagle Gove 22.0 | 56.5
20 One-fourth mile from Wizard Island ---. | 58
20 One-fourth mile off Phantom Ship ---..- 60 |
21 os shore in Eagle Cove. --..-.-.----.-.--- yee |
22 SoA Seal Hee Re Opn sane 56
22 Roncenifesal lake. . 62
22 ..| About 24 miles east of Wizard Island 61
PPA CUT We Se omcid Herne Gl sae techn ce cooeeost eee pecsenss 61

The following intermediate and bottom temperatures were taken, with a Negretti-
Zambra deep-sea thermometer tripped by means of a propeller, such as is used by the
Albatross in her deep-sea temperature work:

| Date. Hour. | Station. Depth. Temper:
1896. | | Feet. | oF.
| Aug. 20) 11 a.m. -| At bottom 4 mile south of Wizard Island = 93 43.5 |
20 | 1.40 p.m At bottom } mile off Phantom Ship- | 866 44
| 22 | 3 p.m About 24 miles east of Wizard Island 555 39
22/4 p.m EeesUG)> 1, 040 41
| 22'| 5 p.m.. At bottom 2

The vertical series taken on August 22, at a station about 24 miles east from the
southeast corner of Wizard Island, proved of very great interest. The surface tem-
perature was 61°; at 555 feet it was 39°; at 1,040 feet 41°; and at 1,623 feet, which
was at the bottom, 46°. In all other American lakes, so far as known, the coldest water
in summer is always at the bottom. The effect of the sun in heating the water of lakes
does not ordinarily reach to any great depth. Observations recently made upon Lake
Champlain by Prof. George C. Whipple and our own observations made in 1896 upon
Alturas and Wallowa lakes showed that the sun’s heat did not much affect the
temperature of the water beyond a depth of 100 feet.

If there be no error in the above observations, it seems that the waters of Crater
Lake are still receiving heat from the rock upon which they rest. The heat of the old
volcano has not entirely disappeared. The coldest water is neither at the surface nor
at the bottom, but at some intermediate depth. The results of these observations are
so unexpected, and the indicated conditions are so unusual, that the matter should
receive further attention. The only possible source of error which has yet suggested

F. C. B. 1897—5

66 BULLETIN OF THE UNITED STATES FISH COMMISSION.

itself is that the propeller may not always have worked properly. It is possible(though
improbable) that in some cases, when beginning to haul up the thermometer the
propeller failed to reverse until some moments iater, in which case the temperature
reading would be for some depth other than the one desired. The observations should
be carefully repeated before the conclusions suggested are accepted.

During the trip of the Mazama party from Ashland to Crater Lake and return a
good many temperature observations were made. All of these possess a certain
value and it has been thought proper to record them in this connection. The air
temperatures were all in the shade unless otherwise stated. The instrument used was
a Wilder protected thermometer.

Miscellaneous temperatures recorded on the Crater Lake trip, August 15 to 26, 1896.

Date. Hour. Station. Air. | Water.|| Date. | Honr. | Station. Air. | Water.
| ee ee = z et
| OF. | oF. | SPE ||5 Weta gh
Aug.13 | 9.10p.m. | Hunt's ranch, 4,400 . | Creek at Crane Creek |------ 41
| feet elevation. camp.
14} 5.00 a.m. SUV sass ssoscsscos: Air at Crane Creek | 36 |........
14 | 9.10 p.m. Camp at Lake of the | _ camp. |
| Woods. Wood River at bridge-|.-..-- 45
#5) (5.00;a. om. |2=2-- do see cee esteeee ce . | Camp 23 miles from |.....- 43
15 | 12.18 p.m. | Greek on way to Mount|------ 54 | Crater Lake, 6,100
| | Pitt. | | feet.
15 | 2.00p.m. | Spring on creek on |....-- | 42.75 19 | 5.25.a.m. |-.-.- Geese sees eo- sa = 49 39
| | way to Mount Pitt. | 19 1.20p.m. | Camp Mazama,Crater| 66 |.-....--.
16) 4.00 a.m. Gearon side of Mount 50} -< oee | Lake, 7,185 feet.
Pit | | 19 | 3.00 p.m. | Small stream on trail |..--.. 45
16 9.42 a.m Surat of Mount Pitt, GBH bso-asse down to lake.
| 9,760 feet. | 19 | 11.00 p.m. | Camp Mazama......-. Ga bee
16 | 11.30a.m. | Summitof Mount Pitt, S34 eee oats | 20 | 6.00.a.m. |.-... ND Sas Sy 25a) Ge) 0 boesseck
9,760 feet, in sun. 20 |12.00m. |..-.. do .. S22 (67.5) |Seeeme re
17 | 4.40 a.m. | Camp returning from BYP | waco ace 20 10.00 p.m. |-.---.- do -- 3) RA Sasso
Mount Pitt. I 21 | 6.00 a.m. |---..- do .-- Ta eee
17 | 12.30 p.m. | Large spring at Peli- |.----.-. 43 |] 21 | 8.45 a. m. |---.- do .-. {G83 |sSsso55-
| can Bay. | 21 | 2.30/p.m. |----- do SCL acces
Ue Reet Gy soos Small spring at Peli- |.--.-- 43.5 21 | 8.00 p.m. |---.- do 58.5 |.--..2.-
ean Bay. 21 | 10.20 p.m. }---.- do 4) 4) | peeosee
Lif ee Ole Creek 100 yards below |.----- | 44.75 | 22) 6.00 a.m. |--... do DD 2 eee
spring at Pelican | | 22} 9.00 a. m. |..-.. do SIRGB ie | Seeeenen
ay. | | 22 4.00 p.m. |..... do .. at essen
17h osetia anise Creek 400 yards below |------ 45.5 23 | 7.35 a.m. |- GD see sSecns eda hitne | | bsSssece
spring at Pelican 23 | 9.00 a.m. Camp Mazama, Cathe- | 66 35.5
Bay. | | dral Spring.
iff ese GY eas Creek at entrance of |.-.--- 47 | 23 | 12.20 p.m. | Camp Mazama.....-... GARG RSE oer
Pelican Bay. | 23 | 3.00 p.m. |..... Gl) =-se-55 Be0s0) ones
UW) Leen anes | Air in shade at Peli- (2 ssneces|| 23 | 9.00 p.m. |.--.- do {COO ee) eee
| can Bay. | \| 24] 5.00a.m. |--... do02ss 226 See ly tae Pe |
UCT PBS ia riisy Ee eS Gh) eo5--Sasecescace BOs earcemeeae 24 | 9.00 p.m. | Rogue River camp, | 61 |-.---.---
17 | 8.00p.m. | Crane Creek camp, SPY| SS aReeine 2,300 feet elevation.
3,900 feet. | || 25 | 5.50 a.m. Rog ue River camp....| 46 |.-.-....
17 | 10.20 p.m. |....- (BO) acti ns occe ns 44 |....--.. | 25 | 9.00 p.m. | Jackson’sranch camp, Gy Whe -ose35
AWA SES Cyisoege Creek at Crane Creek |.-..--. 41.5 |} | _ 1,300 feet.
camp. | 26 | 6.00a.m. | Jackson’sranchcamp.| 57 |.--...--

Life of Crater Lake.—Crater Lake contains no fishes; lakes without inlets or
outlets seldom or never do, for fishes naturally get into one body of water only by
swimming to it from some other body of water with which it is connected. Fishes
never appear de novo in any lake or stream; if they appear there at all it is because
they can swim there from some other lake or stream. Breaks in water continuity, or
even considerable falls, are absolute barriers beyond which fishes can not go. So with
Crater Lake; it has never had any inlet or visible outlet connecting it with any other
body of water in which fishes are found and through which fishes might reach it.

Mammals and birds excepted, only 3 species of vertebrates were found within the
rim of Crater Lake, viz, a snake, a frog, and a salamander.

The snake is the small-headed striped snake, Thamnophis leptocephalus (Baird &
Girard). Two specimens were obtained on Wizard Island. It was not seen elsewhere.

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 67

One frog, Rana aurora Baird & Girard, was also obtained on Wizard Island.
Others were seen about the spring below the camp. The salamander is Amblystoma
macrodactylum Baird. Itis by far the most abundant and most interesting vertebrate
occurring at Crater Lake. Our first specimens were found on the shore under Red-
cloud Cliffs, where we found it to be exceedingly abundant. It was afterwards found
in considerable numbers along the shore in Eagle Cove and a few were obtained about
Wizard Island. More than a hundred specimens were collected and many more
could have been obtained. The majority of the individuals seen were adults, only 4
or 5 still retaining the external gills were seen. These salamanders were found under
the rocks just above the edge of the water of the lake. Sometimes as many as a
dozen or fifteen were found under a single flat stone. These three species are not
uncommon throughout western Oregon and Washington.

Invertebrate life was found to be fairly abundant, and a few plants were found.
By using fine-meshed surface towing nets considerable collections were made. The
following is a list of the species obtained:

PLANTS. ANIMALS—Continued.
Alge:
Nostoc sp.? Colonies common. Insecta—Continued.
Melosira granulata.. Not common. Laccophilus larve. Not common,
Pleurosigma sp.? Several specimens. Deronectes striatellus. Common.
Surirella sp.? Not common. Dytiscus sp.? Not rare.
Vermes:
ANIMALS. | Gordius sp.? One specimen obtained.
Crustacea : | A species of leech. Not common.
Cyclops albidus. Not common. Mollusea :
Cyclops serrulatus. Not common. | Physa sp.? Not common.
Daphnia pulex pulicaria. Very abundant. | Batrachia:
Allorchestes dentata. Not common. Amblystoma macrodactylum.
Insecta : Rana aurora.
Chironomus sp.? Larvie. Very common. Reptilia :
Ephemerid larvee. Common. | Thamnophis leptocephalus.

Limnophilus or caddis fly. Very abundant.

These crustaceans and insect larvie are all excellent food for trout, and were found
in considerable abundance, particularly in Eagle Cove. The small entomostracan,
Daphnia pulex pulicaria, was the most abundant species, and large numbers could be
seen during favorable afternoons swimming at the surface in Eagle Cove. The three
other species of crustaceans seemed far less abundant.

Caddis-fly cases were very abundant on the under sides of rocks lying in the edge
of the water. A small black leech was quite common on the rocks in Eagle Cove and
about Wizard Island. The small gastropod (Physa) was found about Wizard Island
in limited numbers, a species of water-beetle (Dytiscus) was pretty common, and a
single specimen of Gordius, or hair snake, was found near the shore of the island.

There are no water-plants of any size in the lake. On the rocks about Wizard
Island the gelatinous masses of colonies of Nostoe were common. A number of speci-
mens of the diatom, Plewrosigma, were obtained in the towing net, and a few specimens
of two other algze were secured.

Summing up the matter, it may be said that while the conditions at Crater Lake
are not the most favorable to fish-life, there seems to be no reason why trout in limited
numbers might not thrive init. The water is all that could be desired as to purity
and temperature, but the depth is so uniformly great that only small areas of bottom
suitable for spawning-beds are found. The entire absence of all other fish-life and the

68 BULLETIN OF THE UNITED STATES FISH COMMISSION.

very limited vegetation supported by the lake reduce the food supply almost wholly to
small free-swimming crustaceans and insect larve. Both are present in considerable
abundance, but probably not in sufficient quantities to support a large number of fishes.
On the other hand, fishes planted in this lake will have no rapacious fishes with which
to contend; the struggle wiil, therefore, be wholly with the physical environment and
the food supply. It has, therefore, been recommended that a plant of trout be made in
Crater Lake, and it is hoped that the Mazamas who visit the lake hereafter may make
observations to determine the result of the experiment. The best species to plant is
the black-speckled trout of Lake Tahoe, Salmo mykiss henshavci.

LIST OF FISHES.

In the following list of fishes we give only those species of which specimens were
obtained by us during the investigations of 1896. In the nomenclature and sequence
of species we follow Jordan & Evermann’s Check-List,* recently published:

1. Entosphenus tridentatus (Gairdner). Three-toothed Lamprey.

A specimen, 18 inches long, was obtained July 23 in Alturas Inlet, 1 mile above the lake, and
several larvie were dug out of the sand at the head of the lake July 25. Comparing the large specimen
with one 24 inches long, obtained by Mr. Williams at Big Payette Lake, some important differences
are noted. In the Payette example the dorsal fins are separated by a space equal to one-third the
length of the base of the anterior fin, while the Alturas example has the dorsals scarcely separated.
The fins are also much higher in this specimen, the height of the anterior dorsal being contained
31 times in head, measured to first gill-opening, or 1 times in the height of thesecond. The infraoral
lamina has 5 cusps; the supraoral has 3, the middle one being much smaller than the other two; the
buccal teeth before the mouth are unicuspid; on each side of the oral opening are 4 teeth, the anterior
and posterior ones of which are bicuspid, while the other two are tricuspid. The arrangement of the
teeth does not differ materially from that in the Payette specimen. The larve range from 12 to 3}
inches in total length. In all of these the dorsal fin is continuous, though deeply incised.

An example 103 inches long was found attached to a chub in Upper Klamath Lake, November 10.

2. Acipenser medirostris Ayres. Green Sturgeon.

Common near the mouth of Siuslaw River; several taken in a seine near Acme in September,
measuring about 18 inches each in total length. At Gardiner, near the mouth of the Umpqua, a good
many large sturgeon, probably 4. transmontanus, were taken in November and December.

3. Pantosteus jordani Evermann. Western Black Sucker.
During the season’s work this sucker wes found only in Wallowa Lake, Oregon. Two small
specimens were obtained in the upper end of the lake August 24.

4. Catostomus tsiltcoosensis, new species.

Type No. 48479, U.S. Nat. Mus. ; cotypes No. 38, U.S. F. C., and No. 5703, L. S. Jr. Univ. Mus.

Type locality: Tsiltcoos Lake, Lane County, Oregon, where numerous specimens were obtained
December 2, 1896, by Dr. Meek.

Apparently most closely related to C. occidentalis Ayres.

Head 44; depth 5; eye 6} in head; snout 2; D. 13; A. 7; scales 13-65-8, 34 before the dorsal.
Pectoral 1} in head; longest dorsal ray 13; base of dorsal 1?; longest anal ray 1}; ventral 13. Body
rather slender, subterete; head small, snout long and pointed; mouth inferior, overhung by the pro-
jecting snout; lips rather thin, one row of large papillee on upper lip, and about 2 irregular rows of
smaller ones behind or inside of it; lower lip incised nearly to base, 1 or 2 rows of small papille
across the isthmus; lobes of lower lip moderately long and thin, the bases with papillie merging
into plications toward the tips. Eye quite small, the anterior edge of orbit at middle of head. Top
of head flat or very slightly convex between the eyes. Fins small; pectorals short and rounded;

*A Check-List of the Fishes and Fish-like Vertebrates of North and Middle America, in Report
U.S. F. C. for 1895 (December 28, 1896), 207-590.

SALMON INVESTIGATIONS IN COLUMBIA RIVER BASIN IN 1896. 69

ventrals short, rounded, the middle rays but little longer than the others; anal small, somewhat
pointed; margin of dorsal slightly concave; caudal lunate, not deeply forked. Muciferous canals on
head not strongly developed. Scales moderately large; lateral line nearly straight, not running
upward toward nape.

Length of type, 8 inches to base of caudal.

The cotypes agree closely with the type in all characters of importance.

We have compared our specimens of this species with specimens of C. occidentalis (Fig. 2) from
the Sacramento River and find important differences. Comparing examples of the same size, we find
that C. tsilicoosensis has a smaller head, longer, more pointed snout, smaller eye, larger scales, and
much smaller fins; the pectorals in C. occidentalis are faleate while in this species they are more
rounded, and the ventrals also are less pointed.

An example of C. occidentalis 9 inches long has the head 4}; depth 5; eye 5}; snout 2}; D.12; A.7;
scales 13-70-10, 41 before the dorsal.

This species is abundant in Tsiltcoos Lake and Siuslaw River. The specimens from Siuslaw
River were obtained in brackish water.

nee ey

ey,
“9

ry
Re
‘

ER
oe
yy RNY sateen

Fic. 2.—Catostomus occidentalis Ayres.

5. Catostomus macrocheilus Girard. Columbia River Sucker; ‘Yellow Sucker.”

Abundant throughout entire Columbia River basin; obtained this year from Alturas and Wallowa
lakes, in both of which it is abundant. It seems not to occur in Siuslaw River, nor in lakes of that part
of Oregon. Five specimens from Lake Union near Seattle, and one from Lake Washington at Seattle.

6. Catostomus snyderi Gilbert.

One obtained from an Indian at the mouth of Williamson River, November 5. This species was
caught, along with Chasmistes copei and Chasmistes stomias, in the small trap or basket nets which the
Indians were using at that place. Three small specimens, 22 to 4% inches long, were taken in the
seine at the south end of Upper Klamath Lake November 10, and two others, 2 and 24 inches long,
respectively, in a slough at Pelican Bay November 3. No large examples seen in shallow water.

70 BULLETIN OF THE UNITED STATES FISH COMMISSION.

The Klamath Lake Catostomus has been identified, by all who have had occasion to mention it, with
C. labiatus Ayres, upon the supposition that the type of Ayres’s species came from Klamath Lake. But
Ayres plainly states* that it came from Stockton, California. As only one species of this genus is
found in San Joaquin River, C. labiatus becomes a synonym of C, occidentalis, and the Klamath Lake
species is left without a name, a fact first pointed out by Mr. J. O. Snyder, for whom the fish has been
recently named by Dr. Gilbert.t

The type specimen upon which Dr. Gilbert based his description is less than 8 inches long. As the
one which we have from the mouth of Williamson River is much larger (length, 144 inches) and shows
some slight differences, we describe it fully as follows: Head 4}; depth 4; eye 6} in head; snout 24;
maxillary 3}; mandible 2}; interorbital 2}; width of mouth 3} in head, more than half length of snout ;
greatest width of lower lip ? diameter of eye; D. ii, 11; A.7; scales 13-70-11. Body rather slender; head
long, mouth moderate, horizontal; lips thick papillose, the upper with about 4 or 5 rows of papille,
lower with about 7; lower lip divided nearly to base, leaving only one row of papillie crossing the
symphysis; premaxillary not much projecting and not forming a prominent hump; mavxillary rather
short, not reaching vertical at front of anterior nostril; eye equally distant between snout and poste-
rior edge of opercle; mucous canals on head forming raised ridges, the pores conspicuous. Fins mod-
erate; origin of dorsal a little nearer snout than base of caudal, sixth spine over insertion of ventral;
pectoral 1} in head, reaching slightly more than two-thirds distance to ventrals; ventrals not quite
reaching vent, the seventh ray longest, 1% in head; anal long, pointed, reaching to base of caudal, 1}
in head. Seales crowded anteriorly, about 32 transverse rows in front of dorsal, strongly ridged, the
margins crenate.

7. Chasmistes stomias Gilbert.

One specimen, 15 inches long, obtained from an Indian at Klamath Falls, November 11. Head 44;
depth 44; eye 7; snout 2*; maxillary (measured from free end to tip of snout) 34; mandible 24; D. ii,
11; A.i,7; scales 13-85-10; interorbital width 24; vertical depth of head at mandibular articulation
2). Head small, body heavy forward, the back strongly and regularly arched from snout to origin of
dorsal fin, thence declined in a nearly straight line to base of caudal; ventral surface nearly straight.
Premaxillary spines strongly protruding, forming a prominently projecting snout; mouth rather small,
inclined upward at an angle of about 40°, maxillary scarcely reaching vertical trom front of anterior
nostril; width of mouth 12 in snout or 4? in head; upper lip thin, without papillie; lower lip thin,
interrupted at the symphysis, forming narrow lateral lobes the width of which is about 2? times in
their length; faint indications of a few papillx ; mucous canals forming ridges, the pores conspicuous ;
gill-rakers long, narrowly triangular at the tip when viewed from behind, densely tufted on the
anterior edge; fontanelle narrow, its length 2? in the snout, its width about + its length. Fins all
large; the origin of the dorsal a little nearer tip of snout than base of caudal, the sixth ray over base
of ventral, its base 1? in head, the free edge nearly straight, the last ray 13 in the first, which is 1}in
head; pectorals scarcely faleate, reaching a little more than two-thirds distance to base of ventrals,
their length 1} in head; ventrals long, reaching vent, the rays gradually increasing in length from the
outer to the seventh and eighth, which are longest, the ninth and tenth being but slightly shorter,
the leneth of the longest ray 1% in head or about } longer than the first; anal long and pointed, the
fourth ray longest, reaching base of caudal, 1} in head; each ray of anal fin with 8 to 12 strong
tubercles; candal lobes about equal, their length 1{ times the middle ray.

This specimen agrees with Dr. Gilbert’s type,} with which we have compared it. In the type the
mouth is rather more oblique, the maxillary is slightly longer, and the anal fin is longer.

8. Chasmistes copei, new species. Klamath Indian name ‘‘ Tswam,”

Type No. 48224, U. S. N. M. (collectors’ No. 871), a specimen 16 inches long. Type locality:
Northwest part of Pelican Bay, Upper Klamath Lake, Oregon; collected (in gill net) November 6, 1896,
by Messrs. 8. E. Meek and A. B. Alexander.

Head 33; depth 4; eye 6}; snout 2}; maxillary (measured from free end to tip of snout) 3;
mandible 22; D. ii, 10; A. i, 7; scales 13-80-12; interorbital width 2}; vertical depth of head at man-
dibular articulation 24. Head large, cheek very deep, the depth equal to distance from tip of snout
to nostril; body stout, back scarcely elevated, caudal peduncle rather short and stout; ventral surface
somewhat convex. Premaxillary spines less protruding than in C. stomias, not forming a prominent

*Proc. Cal. Ac. Nat. Sci., 1, 1855, 33.
tBull. U.S. F.C. 1897, 3.
{ Described in Bull. U.S. F. C. 1897, 5, with figure.

SALMON INVESTIGATIONS IN THE COLUMBIA RIVER BASIN IN 1896. ol:

hump; mouth large, inclined upward at an angle of 45°, maxillary not nearly reaching vertical at front
of anterior nostril; width of mouth 1} in snout, or 4 in head; upper lip thin, without papillie; lower
lip thin, entirely without papille, interrupted at symphysis, forming rather broad lateral lobes; pores
on head very conspicuous; gillrakers larger than in C. stomias, broadly triangular at tip when viewed
from behind, densely tufted on anterior edge, each appendage more or less bifid and club shaped, closely
resembling those of C. liorus; fontanelle narrow, its length 2} in snout, width one-fifth its length.
Fins all small; origin of dorsal a little nearer snout than base of caudal, its sixth ray over base of
ventrals, free edge straight, base 2} in head, last ray a little less than 2 in first, which is 2 in head;
pectorals somewhat falcate, reaching slightly more than half distance to ventrals, their length 1% in
head; ventrals very short, reaching only two-thirds distance to vent, free end nearly straight; outer
ray longest, 27 in head; inner shortest, 34 in head; anal fin short, bluntly pointed, not reaching base of
candal, third and fourthrays longest, 1{in head; no tubercles on anal rays; caudal lobes equal, length
about 1 times the middle ray. Scales small and crowded anteriorly, about 14 rows downward and
backward from front of dorsal to lateral line, 11 vertically upward from base of ventral to lateral
line, about 38 oblique series before dorsal; lateral line nearly straight, with about 80 scales. Entire
upper parts of head and body, and sides nearly to level of base of pectorals, dark olivaceous; under parts
abruptly whitish or yellowish in alcohol; a dark spot in upper part of axil; dorsal and caudal dark;
pectorals dark on inner surface; ventrals and anal plain.

Fic. 3.—Chasmistes copei Evermann & Meek, new species.

From Chasmistes stomias this species is readily distinguished by its larger head, larger, more
oblique mouth, less prominent snout, and very small fins. The differences in the fins are very great,
particularly in the ventrals, as may be seen in the accompanying illustrations. It differs from C.
brevirostris, as characterized by Dr. Gilbert, in its much larger, more oblique mouth, the absence of
papillie on the lips, and shorter fins.

We name this species for the late Prof. Edward Drinker Cope, who wrote the first paper on the
fishes of Upper Klamath Lake.

Six nominal species of suckers have thus far been described from the Klamath Lakes, viz: Chas-
mistes luxatus and Chasmistes brevirostris by Cope in 1879; Catostomus rex by Rosa Smith Eigenmann in
1891; Catostomus snyderi and Chasmistes stomias by Gilbert, and Chasmistes copei by Evermann &
Meek, the last three in the present Bulletin. Mr. A. Seale has recently taken C. luxatus as the type
of his new genus Deltistes, which he bases upon the peculiar structure of the gillrakers. Dr. Gilbert
finds that Catostomus rex is a synonym of Deltistes luxatus. As now understood, we therefore know
from Upper Klamath Lake one species of Catostomus, one of Deltistes, and three of Chasmistes.

9. Mylocheilus caurinus (Richardson). Columbia Chub.

Obtained from Little White Salmon River, Lakes Washington and Sammamish at Seattle, and
Lake Pend d’Oreille at Hope, Idaho. Abundant early in September in lagoon at mouth of Little White
Salmon River, and seen in considerable numbers at steamboat landings between Portland and The
Dalles. It does not occur in the Redfish Lakes in Idaho, nor has it been recorded from Wallowa
Lake, nor from any of the streams or lakes south of the mouth of the Columbia.

ri? BULLETIN OF THE UNITED STATES FISH COMMISSION.

10. Ptychocheilus oregonensis (Richardson). Squawfish; Yellowbelly.

This large minnow is abundant throughout the Columbia River basin except the portion above
Shoshone Falls. In Montana, and perhaps elsewhere, it is called “squawfish”; at Sawtooth, Idaho, it
is known only as “‘yellowbelly”; at Mapleton, Oregon, ‘“‘chub” was the only name heard applied to
it; while at Tsiltcoos Lake it was called ‘‘dace.” Specimens obtained from Lakes Alturas, Pend
d@ Oreille, Gamlin, Washington, Tsiltcoos, Whoahink, and Tahkenitch, and from Siuslaw River. One
was taken on a set line in Whoahink Lake, at a depth of 78 feet, on December 3.

Scale and fin formule and comparative measurements of specimens of P. oregonensis from Lake Washington.

Length | | | | Length :
in Head. Depth. Eye. | Snont. ee Anal. Scales. in | Head. Depth. Eye. | Snout. Dor Anal. Seales.
| inches. | | | | hire || inches. | | | | shes |
| | | | |
7.25..| 3. 64 | 5.33 | 5.60 | 1.80 10 | 9 76 || 5.25 | 3.78 5. 00 | 5. 00 | 1. 42 10 | 9 | 77
7.75..| 3.67 5.00 | 5.80 2. 00 10 | 9 80 || 5.00. 3.60 | 4.86 | 5.25 | 1.67 | 10} 9 | 74
7.00..| 3.50 | 5.00 | 5. 60 1.80 10 9 77 || 4.63... 3.83 | 4.75 | 5.00 1. 67 10 9 76
5. 75..| 3. 57 | 5.00 | 5.17 1. 67 10 | 9 74 || 4.38...) 3.83 4.67 | 4.67 1.50 10 9 | 74
5.88..| 3.44 |} 5.00 | 5.00 1.50 10 | “) 74 | 4.38...) 3.67) 4.67 5.00 1. 50 10 9 77
5.50--| 3.67 | 5.00| 5.00] 1.50] 10| 9 75 || 3.17-. | 3.60 | 4.60 | 4. 60 1.40] 10| 9] 78
| | \ | | | |

11. Leuciscus bicolor (Girard).

Judging from our collections, this is the most abundant minnow in Upper Klamath Lake. About
60 were obtained at the lower end of the lake November 10, and over 100 from a small creek at Pelican
Bay November 5. At the time of collecting this species was found in the shallow water of the little
creeks and passages among the tules. The largest are 9 to 10 inches long, but only a few exceed 3
inches. Examples 2 to 3 inches long gave the following measurements: Head 4; depth 5; eye 3} to
32; snout 4? to 44; scales 59; teeth 2, 5-5, 2. See illustration on page 7 of this Bulletin.

LEE
gas:
aaah

ae aa 5
qari

dads

=

Fic, 4.—Leuciscus siuslawi Evermann & Meek, new species.

12. Leuciscus siuslawi, new species.

Type No. 48480, U. S. N. M., a specimen 5 inches long. Cotypes No. 433 U. S. F. C.; No. 48231,
U.S.N.M., and No. 5702, L. S. Jr., Univ. Mus. Type locality: Siuslaw River at Mapleton, Oregon.
Collected September 8, 1896, by S. E. Meek.

Head 41; depth 4}; eye 4; snout 31; maxillary 32; D. ii, 9; A.ii,12 or 13; scales 11-58-8; teeth
2, 4-5, 2, somewhat hooked.

Body rather slender, slightly elevated and somewhat compressed; head small and pointed, cheek
not deep; snout pointed, somewhat longer than eye; mouth moderate, somewhat oblique, maxillary
just reaching vertical at front of orbit; jaws subequal, the lower sometimes slightly projecting; eye
large, not as great as snout. Origin of dorsal fin behind base of ventrals and much nearer base of
caudal than tip of snout, the longest ray 14 in head, greater than base of fin; origin of anal fin under

SALMON INVESTIGATIONS IN THE COLUMBIA RIVER BASIN IN 1896. 73

last dorsal ray but two, its height equal to that of dorsal, its base equal to its longest ray; free edges
of dorsal and anal nearly straight; pectoral 1} to 14 in head, not reaching insertion of ventrals;
ventrals short, 14 in pectoral, reaching anus; caudal deeply forked; lateral line complete, decurved;
color in spirits, brownish or olivaceous above, middle of side with a broad dark band involving the
lateral line anteriorly and posteriorly, but lying chiefly above it mesially; middle of side from gill
opening to beneath dorsal fin with a broad rosy band, following closely beneath the lateral line;
lower part of sides and under parts silvery, dusted over with fine dark specks; a light yellowish band
extending backward from upper posterior border of eye nearly halfway to origin of dorsal fin; cheek
with a silvery or golden crescent; top of head dark; opercles dusky silvery; snout dusky; fins plain,
dorsal and caudal somewhat dusky. Numerous specimens were obtained in the Siuslaw River and
one from Tsiltcoos Lake.

At Mapleton both this minnow and Ptychocheilus oregonensis are known as ‘‘chub,” while on the
lakes south of Florence they are called ‘‘dace.” It is probably common in these lakes, but a single
specimen only was obtained there.

Compared with Leuciscus balieatus, which our specimens most closely resemble, the latter are seen
to have smaller anal and dorsal fins, a more slender body, smaller and more slender head, and longer,
more pointed snout. The extent of variation in proportional measurements and in te number of anal
fin rays appears to be much less than in L. balteatus.

This species also resembles Leuciscus cooperi of Girard. In the type of L. cooperi (No. 238,
U.S. N.M.) the lower jaw is notably shorter and the snout more pointed.

In the following statement will be found measurements of a number of specimens from Siuslaw
River, Mapleton, Oregon:

|

| || 3
hana Head. Depth. Eye. |Snout. eal, Anal Seales. [ett ea, Depth. Eye. Snout. pal, Anal. Seales.
| = - = | | ———- | — —— =
|
Inches. | | | || Inches. | | |
5.75 4.20 4.17 | 4.00 3.50 9 13 | 11-58-7 || 4.50} 4.00) 3.83 | 3.50 3.50 9 12 | 10-60-7
5.50 | 4.00 4.00 | 4,00 | 3.60 8 11 | 12-59-7 || 4.50 | 4.00| 4.00 | 3.80| 3.50 9 12 | 10-59-7
5.50 | 4.00 4.17 | 4.00 3. 60 9 13 | 10-59-7 4.50 | 4.00 4.40 3.40 3. 80 9 12 | 10-61-7 |
5.50| 4.17) 4.17) 4.00] 3.60| 9| 13) 12-646 || 4.50| 4.00] 3.80| 3.80] 3.75] 9| 413 | 10-59-7
5.25 | 4.00) 4.00 | 4.00) 3.75 8 12) 12-60-7 || 4.25) 4.00] 4.00 | 3.60) 3.80 9 11 | 10-63-6
5.00] 4.00] 3.80) 3.50| 3.60} 8] 13) 12-60-7 || 4.00) 4.00] 4.00|3.20) 4.00] 9| 13 | 10-61-7
4.50] 4.00) 4.17 3.50) 3.50 9 | 13. 12-64-6 )
| | if

The anal fin rays were counted in 16 additional specimens with the following results, the average
for the total 29 specimens being 12}:

r, » |
Total No. sf ||. Total No. a Total Fle: of
length. | 202" | length. ana’ |) length. te
rays. || rays. | rays.
| |
Inches. | | Inches. Inches. |
5.75 11 2.50 | 13 2. 63 13
5.75 13 2. 25 12 2. 63 13
12 || 2.25 | 2 2. 63 13
13 || 2.81 1 || 2.50 14
1 || 2.75 3 |} 2.13 13
Reccessbes mates bee eee ae bee Res) iS} 12
| Il

a From the outlet of Tsiltcoos Lake.

13. Leuciscus balteatus (Richardson).

Abundant throughout the Columbia River basin except above Shoshone Falls. Specimensobtained
from Alturas, Pend d’Oreille, Gamlin, and Washington lakes. The variation in the number of anal fin
Tays of specimens from these localities is shown in the following table. The length is measured from
tip of snout to base of caudal fin. The last ray, though usually deeply divided, is counted as one.
There are usually two rudimentary rays at front of fin, which are not counted, Of 30 specimens
from Gamlin Lake, 1 has 13 rays, 7 have 15, 14 have 16, 4 have 17, and 4 have 18, the average being 16
rays. Of 10 specimens from Lake Washington 5 have 14 rays, 3 have 15, 1 has 16, and 1 has 18, the
average being 15 rays.

74 BULLETIN OF THE UNITED STATES FISH COMMISSION.

Table showing variation in the number of rays in the anal fin in Leuciscus balteatus.

Gamlir Lake. Lake Wash-
ington.
Length| Length | Length | || Length
in Anal. in Anal.|| in |Anal.) in Anal.
inches. inches. | || inches. | inches.
|
1.88 | 16 || 1.63 | 16 |) 1.75 | 16 | 2.75 | 15
1.88 | 16 1.50 | 16 || 1.75 | 1g || 3.13 | 14
1.88 | 16 || 2.00 TRIN aS ay) | 2.50 14
TeCEY |) call 2.13 | 15 || 2.00 | 16 3.13 | 15
1.75 | 15 || £75 | a7 || 200 | 16 |) 313°] 14
2.00 16 1.88 | 16 || 1.63 | 18 8.50 | 14 |
1.88 | 15 ASSGE A183 Tees Nae 3.50 | 16
1375) We aoe |) 15 | 1.75 | 13 4.00 | 15
Te hath be Eger ity GGL) i) Tape Chee | ie
1.63 | 16 |) 1.75 | 15 | a4.75 | 16 3.88 | 14

a From Lake Pend d’Oreille.

14. Rutilus bicolor (Girard).

Numerous specimens obtained in a slough at Pelican Bay and others at the south end of Upper
Klamath Lake. Some were caught in a gill net, others in a small dip net, such as is used by Klamath
Indians; a few were taken on a set line, but the majority were obtained by means of a small seine, in
small sloughs or lagoons at the south end of Upper Klamath Lake. It seems to be one of the most
abundant species occurring in the lake. The largest measure about 9 inches in total length. There
does not appear to be much variation in the number of scales or fin rays, or in proportional measure-
ments, as is shown in the following table. The Indians use this species for bait when fishing for
trout near Klamath Falls, the average length of those thus used being about 4 inches. It is caught
in large numbers in dip nets in the edge of the tules. This species is figured on page8 of this Bulletin.

Table showing variation in 13 specimens of Rutilus bicolor from Upper Klamath Lake, Oregon.

* at |
Tene Head. Depth. Eye. | Snout. Mal: Dorsal.| Anal. Seales.
9. 00 3.75 3.75 7. 00 3. 50 4.00 9 8 | 49
9. 25 3.75 3. 67 6. 00 3,50 3. 75 9 8 49
6. 00 3.75 3.75 5. 50 3.75 4.25 9 8 48
4.50 4. 00 4.00 4.50 4.00 4. 20 9 8 47
4.50 3.75 3. 75 4. 50 4.00 | 4.25 9 8 48
4. 25 4.00 4.00 4.50 4.00 | 4.25 9 Sh 46
5. 25 3. 50 4.00 5. 00 3. 50 4.33 9 8 48
3.75 3.50 4.00 4.25 3, 67 4.25 9 8 49
3.75 3. 50 4.00 4.50 4.00 | 4. 00 9 8 44
3. 50 3. 50 3. 75 4.25 4.00 | 4.50 9 8 | 46
3. 50 3. 50 4.25 4, 20 4.00 4.33 8 7 47
3. 50 3. 67 4.25 4.00 4.00 4.25 9 7 47
2.50 3. 60 4.33 | 3.50 4. 50 4.25 9 8 | 46
| |

15. Rhinichthys dulcis (Girard).

One specimen from Wallowa Lake. The origin of the dorsal is midway between tip of snout and
base of caudal peduncle,

16. Agosia klamathensis, new species

Type No. 48225, U. S. N. M.; cotypes No. 5704, L. S. Jr. Univ. Mus. and No. 451, U. S. F. C.
Type locality at mouth of the small creek which flows into the arm of Upper Klamath Lake called
Pelican Bay, where numerous specimens were collected November 3, 1896, by Dr. Seth E. Meek and
Mr. A. B. Alexander. Length of type 24 inches, measured to last caudal vertebra.

Close to Agosia nubila.

An examination of a large series of specimens of Agosia from Upper Klamath Lake and a comparison
of them with specimens of 4. nubila from various localities show the fish from this lake to possess certain
characters which serve to distinguish it from typical 4.nubila. While the differences are slight, they
are plainly evident and must be recognized, and we therefore describe this form as a new species.

Head 4; depth 43; eye 44 in head; snout 3}. D.i, 8; A.i, 7; scales about 14-78-10 (average, 73
in 49 specimens).

SALMON INVESTIGATIONS IN THE COLUMBIA RIVER BASIN IN 1896. (8)

Body robust, subeylindrical, back somewhat elevated; snout rather long, mouth inferior, little
oblique, the lower jaw included; maxillary not reaching front of orbit; upper lip without frenum;
barbel present, but small. Lateral line incomplete, interrupted in many places, about 30 pores devel-
oped. Origin of dorsal fin midway between front of pupil and base of caudal fin; pectorals rather
short, reaching about three-fourths the distance to base of ventrals; ventrals reaching vent; anal large,
its longest ray 1} in head; caudal widely forked.

Color in alcohol, olivaceous, mottled and blotched with darker on back and sides; under parts
pale; an obscure pale streak from eye to base of caudal fin, below which is a broad dark band; dorsal,
pectorals, and caudal dusky ; other fins plain; a black blotch at base of caudal.

An examination of the 48 specimens which we designate as cotypes shows all the important char-
acters to be fairly constant. The barbel is in a few cases obscure or possibly absent. The number
of scales in a transverse line varies from 68 to 78, the average being 73. The lateral line is in all cases
incomplete, though in different degrees; sometimes it is continuous for only 6 or 8 scales, after which
there are several interruptions and only 8 to 10 more pores; in others there are 20 to 30 pores in a
continuous series.

9
BER

Pras

seRea ey

Se Daa

Fic. 5.—Agosia klamathensis Evermann & Meek, new species.

This form is distinguished from Agosia nubila chiefly by the smaller size of the scales. This
difference will appear from an examination of the tabular statement which follows:

Table showing number of scales in Agosia nubila and Agosia klamathensis,

| No. of |Variation) Average
Locality. specimens in number numberof,
examined, of scales.| scales.

— — = = = = 3 = |

Agosia nubila:
Colville River, Meyers Falls-......-.-..-.-- 6 52-57 55
Little Spokane River, Dart’s Mill-- E 9 52-68 62
Hangman Creek, Tekoa, Wash-.--. z 10 51-65 57
Lake Coeur d’Alene....-...-- A 9 64-70 66
Newaukum River, Chehalis. aoe (?) 53-58 | 55
Boise River, Caldwell...-.--- ae if 56 | 56
Potlatch Creek, Lewiston. ete 10 60-66 62
Pataha River, Starbuck .--- 9 60-64 62
Walla Walla River, Wallula. 15 47-55 | 49
Mill Creek, Walla Walla. --- 20 52-61 | 55
Umatilla River, Pendleton. . 3 48-58 52
Columbia River, Umatilla.----.-. 7 48-57 | 53
Natchess River, North Yakama -.. - (2) 53-58 56
Shookumechuck River, Chehalis.-..--..--.-- 6 50-57 55

Agosia klamathensis :
Pelican Bay, Upper Klamath Lake...--.---- 49 68-78 73

17. Clupanodon ceruleus (Girard). California Sardine.

Very abundant in Siuslaw River about the cannery wharf at Florence during the canning season.
It seems to disappear as soon as the fall rains set in and the river becomes filled with fresh water.
The specimens seen were each about 2} inches in length.

76 BULLETIN OF THE UNITED STATES FISH COMMISSION.

18. Alosa sapidissima (Wilson). Common Eastern Shad.

During the salmon fishing season of 1896 in the Siuslaw River about a dozen shad were taken—
one in a salmon gill net near Mapleton about the middle of October, the others in salmon seines at
Acme in September and October. The specimen caught near Mapleton is a male 18 inches long and
weighing about 2} pounds. The spermaries are but slightly developed, indicating that the spawning
season would be some months earlier than when this fish was taken. Head 4; depth 3; orbit 44;
snout 47; maxillary 2; gill-rakers 35-++68 on the left and 36-+68 on the right.

The first planting of shad on the Pacifie coast was made in 1871 by Mr. Seth Green, who liberated
about 10,000 fry in the Sacramento River 275 miles above Sacramento. The second plant was made in
July, 1873, by Mr. Livingston Stone, who placed 35,000 fry in the Sacramento River at Tehama. Vari-
ous subsequent plants have been made by the U.S. Fish Commission in California, all at Tehama, the
last plant having been made in 1880. Besides these plants made in the Sacramento River, 60,000 fry
were placed in the Willamette River at Portland and 10,000 in Snake River in 1885. In 1886, 550,000
fry were liberated in the Willamette River at Albany, and 300,000 in the Columbia at Wallula June-
tion. In 1884, 1885, and 1886, 2,651,000 shad fry were placed in the Colorado River at The Needles,
Arizona. No investigation has ever been made for the purpose of determining the result of the plant-
ings made in the Colorado River and it is not known whether any of the fry survived. The results
from the plants made in California and Oregon, however, are little less than maryelous, as shown by
Dr. H. M. Smith in a recent paper.*

From the Sacramento and Columbia rivers shad have spread to Los Angeles County, California,
on the south, and Wrangeil Island, Alaska, on the north. They had spread from San Francisco to the
Columbia River as early as 1880, and by 1882 had been taken at various places along the Washington
coast. The only passages through which the planted shad could reach the sea are the mouth of the
Columbia River and the Golden Gate. From these points they have spread up and down the Pacific
coast a distance of more than 1,300 miles. This is greater than its range in latitude on the Atlantic
coast. Not only have they spread to these distant points, but shad have been reported from a number
of intermediate places, among which are Monterey Bay, Drake Bay, Klamath River, Rogue River,
Umpqua River, Siuslaw River, Fraser River, and the north end of Vancouver Island.

These facts in the distribution of the shad on the Pacific coast are of extreme interest and
importance as bearing directly upon the belief, still more or less prevalent, that anadromous fishes
possess a mysterious geographic instinct which leads them back to the stream in which they were
spawned. While the number of shad entering the Columbia and San Francisco Bay is far greater than
that for all the other rivers combined, it is nevertheless true that many shad have found their way to
other and distant streams. The extent to which this has occurred seems to us fuily sufficient to
disprove the possession by the shad of any ‘special geographic instinct.” We believe the same to
be true of the various kinds of salmon and other anadromous species. The question is in need of further
research, but investigations already made indicate that anadromous fishes, like migrating birds, are
guided in their movements by landmarks or other tangible physical features or conditions.

The young shad hatched in any given stream go down to the sea, but probably do not ordinarily
wander far from the mouth of that particular river. When they become mature, and physiological
unrest, due to the development of the reproductive organs, comes on, they begin to search for suitable
spawning-beds, and the chances are they will find the river in which they were hatched. The
majority will find this stream, while those that wandered farthest from its mouth may find others
which they will enter.

19. Coregonus williamsoni Girard. Rocky Mountain Whitefish.

Common throughout the Columbia River basin. Specimens obtained from Alturas, Pend d’Oreille,
and Wallowa lakes, Des Chutes River at Sherar’s Bridge, and Big White Salmon River. A specimen
10 inches long, taken in the gill net in Alturas Inlet July 27, has the head 5; depth 44; eye 42; snout
33; maxillary 33; D.12; A.11; scales 86.

Another example, a female, 12} inches long and with well-developed roe, was taken with the hook
at Sherar’s Bridge August 30. Head 5; depth 43; eye 5; snout 3}; maxillary 34; D.13; A.11; scales
88. In this specimen the adipose fin is remarkably large, its base being half length of head, its
height 64 in head. This species spawns in October at Big Payette Lake, in Idaho, but the condition
of the ovaries of the specimen taken at Sherar’s Bridge indicates a much earlier spawning period.

* A Review of the History and Results of the Attempts to acclimatize Fish and other Water Animals
in the Pacific States, in Bull. U.S. I. C. 1895, 407.

SALMON INVESTIGATIONS IN THE COLUMBIA RIVER BASIN IN 1896. 77

20. Oncorhynchus keta (Walbaum). Dog Salmon.

According to Mr. Alexander, this salmon usually commences to run in the southern part of Puget
Sound about the middle of October and continues until the first of December. In 1895 and 1896 they
were quite numerous. During fall and winter all the small creeks, lagoons, and sloughs near Duwam-
ish and Cedar rivers are filled with dog salmon, and boys find great amusement killing them with
clubs and stones. In the rivulets by the roadside, where the water is not over 2 or 3 inches deep, dog
salmon may be seen trying to get farther upstream. At such times they are in poor condition and no
use is made of them.

The condition of the dog salmon in Jannary, 1897, was unusually good. These fish were plump in
appearance and marked with that brightness which they possess when first coming from the ocean.
In the opinion of some of the dealers many of the January (1897) run were fresh from the ocean. Their
eggs, like those of the steelhead, showed various degrees of development; most of the fish were well
advanced, however. It has been only a few years since it was known that any species other than the
steelhead was to be found in Puget Sound during winter, but it is now thought that salmon have
always been more or less plentiful in Puget Sound during the winter months—not a heavy run, but
enough to supply the local demand. In former years, there being no sale for salmon after the canneries
were closed, fishing was almost wholly suspended until the next season. As soon, however, as the
experiment of shipping fresh salmon to eastern markets proved a success, a new industry was opened,
and fishermen who had hitherto given no thonght to winter fishing now began to investigate the
waters of Puget Sound out of season, and the result is that a winter fishery of considerable importance
has been introduced in Seattle and other places on the sound and is increasing yearly. The steelhead
is the most valuable fish for shipment to eastern markets, as it reaches its destination in better condi-
tion than other species.

21. Oncorhynchus tschawytscha (Walbaum). Chinook salmon.

Very few chinook salmon came to the headwaters of Salmon River in 1896, perhaps not over a
dozen, where there were about 1,000 in 1895, The number which came to the Wallowa spawning-beds
was also very small. Important spawning-beds were found in Little White Salmon River, and con-
siderable numbers were found spawning in Big White Salmon River, Eagle Creek, and Tanner Creek.

Young chinooks were found in Siuslaw River in considerable numbers and a few were obtained in
Lake Washington. Two specimens, 24 and 22 inches in length, respectively, from the mouth of Big
White Salmon agree perfectly with those gotten in 1895 at Alturas Lake. These two specimens were
evidently hatched the preceding winter.

Nine specimens from the Siuslaw River, caught with hook and line at Florence, October 14, 1896;
length, 5, 54, 6, 64, 64, 7,7, 7, and 74 inches respectively. Salmon of this size are very abundant about
the cannery wharf during the canning season. ‘They are easily caught with hooks baited with
salmon eggs.

Two specimens 11} and 114 inches in length from Seattle, Washington, caught December 8, 1896.
These specimens were no doubt hatched during the, winter of 1894-95. One specimen, 6 inches long, was
taken with a seine near the mouth of the Siuslaw River December 9, 1896. It is the same age as the
specimens mentioned above from Florence, One small, nearly ripe male, length 18} inches, from the
Siuslaw River at Mapleton, Oregon, was caught on a hook baited with salmon eggs. A few other
specimens of the same size were caught in the same way while fishing for trout about October 21. We
are informed that it is not uncommon to catch these fish with baited hooks.

A few of these small male salmon were seen on the spawning-beds in the North Fork, near
Minerva, October 23 and 24. They were mutilated the same as the larger ones, and one wasin a dying
condition.

On September 9 Mr. A. B. Alexander examined 129 chinook salmon in the Florence cannery; of
these 76 were females and 53 were males; 25 of the females and 13 of the males were fully developed.
On September 11 he examined 546 chinooks; of these 317 were females and 229 males; 229 females and
110 males were nearly ripe. There were among this number 25 small fish from 18 to 25 inches in
length; these were as fully developed as the large ones.

During the latter part of September and early part of October, Mr. Alexander examined many
chinook salmon at Celilo and obtained much valuable information as to their spawning condition.
In one lot of 119 fish examined 57 were males and 62 females; 34 males and 47 females were nearly or
quite ripe, and would haye spawned by the 8th or 10th of October.

In the following tables is given a record of the fish examined by Mr. Alexander.

78 BULLETIN OF THE UNITED STATES FISH COMMISSION.

Chinook salmon taken in fish-wheels by Mr. I. H. Taffe, Celilo, Oregon, September 18 to 22, 1896.

Males. Females.
Date. j= | Total
= _ | Well de- _. | Well de- | c
Number. veloped. | Number. veloped. |
19 10 9 | 9 28
8 | 5 10 | 6 18
18 11 35 | 26 53
12 | 8 8 | 6 20
57 | 34 62 | 47 119
| |

Among the 119 chinook salmon, 13 small ones were found, all males, and their milt as fully
developed as that in larger fish. Sometimes the wheels take these fish in considerable numbers.

Chinook salmon taken by Mr, F. A. Seufert, at Celilo, Oregon, September 25 to October 13, 1896.

Males. Females.

Date. - (==. Total
Well de- Well de- 2
Number. veloped. Number, | veloped.
a i |
1896.

Sept. 25. ----- --.-<-....---o0- 82 67 64 | 61 146
26 15 11 ilyf 17 32
109 82 89 76 198
56 47 70 63 126
56 50 49 44 | 105
Oct. 60 50 76 73 136
122 105 148 144 270
63 55 67 65 130
43 38 51 51 94
7 69 88 64 165
683 574 719 658 | 1,402

In a total of 683 males and 719 females Mr. Alexander found 574 males and 658 females which, in
his judgment, were nearly ripe. In other words, 84 per cent of the males and 90 per cent of the females
would have spawned between the first and middle of October. These observations indicate that
chinook salmon can be obtained abundantly at Celilo by wheels and that if retained a few days they
would be ripe enough for stripping. Whether the wheels seriously injure the fish can be determined
only after actual experiment with fish so caught. If the wheels do not injure them they can probably
be kept until fully ripe in properly constructed retaining boxes or ponds, as was demonstrated by
experiments at Mapleton, Oregon.

According to Mr. Alexander chinook salmon appear in the lower part of Puget Sound about the
1st of May and continue to increase in numbers until July. Scattering ones are taken, however,
throughout the year in all parts of the sound.

22. Oncorhynchus kisutch (Walbaum). Silver salmon.

Our collection contains specimens of this species as follows:

Three ripe males, one 16 inches in length, caught December 1 in Whoahink Lake, the other two,
16 and 17} inches respectively, December 3, in Tsiltcoos Lake. Specimens of this size and very much
smaller were frequently caught in the seine at Acme. Not less than 50 of these were examined. All
were males with sexual organs as highly developed as in the larger ones.

One large ripe male from Tsiltcoos Lake, caught in a gill net December 3, 1896. Immaculate, back
with a bluish tinge, body bright red, mouth much distorted. Locally called ‘‘hookbilled silverside.”

There is a moderately small form of the silver salmon in the Siuslaw River called blueback,
which resemble the true blueback in size, form, and color, but are more spotted. Two specimens, 8} and
10 inches in length respectively, from Tsiltcoos Lake near outlet, December 2, 1896. Back, brownish
blue; dorsal, nearly black, darker on posterior part; pectorals, light brownish; ventrals, white; anal
fin with a dark shade. These specimens were probably hatched during the winter of 1894-95,

Two specimens from Seattle, Washington, 10} inches in length.

SALMON INVESTIGATIONS IN THE COLUMBIA RIVER BASIN IN 1896. 79

Six mature specimens from Union Lake near Seattle, Washington, 4 males and 2 females.

Of the small specimens, no doubt hatched during the winter of 1895-96, our collection contains
the following:

Two specimens, 2% and 34 inches in length respectively, hatched during the winter of 1895-96 by
Mr. L. E. Bean at Mapleton and retained in a small spring brook until September 9, 1896.

One, 3} inches long, caught in a seine in Siuslaw River below Florence December 8, 1896.

Two specimens, 5} and 52 inches respectively, caught with a seine in Tsiltcoos Lake December 1,
1896. All three with parr marks present; no red on sides; dorsal, adipose, and caudal fins yellowish;
caudal reddish near tips; pectoral and ventrals yellowish.

During the latter part of September and early part of October Mr. Alexander was at Celilo, on the
Columbia River, and made some valuable observations upon the silver salmon. He examined a total of
2,268 fish of this species, all of which he found well advanced, indicating to him that they would
spawn before the middle of October.

Silver salmon examined by Mr. Alexander at Celilo.

Male. | Female. |
OEE Num- | Well de-| Num- | Wellde-| Total.
ber. veloped.) ber. | veloped.
— = \_ | d
Taken in wheels by Mr. Taffe: | |
Sept. 18, 1596 ........... eee 5 2 1 0 6
Sept. 19, 1896 .. 6 4 2 1 8
Sept. 21,1896 .. za 9 7 13 | 8 22
Sept. 22,1896 ....-.-. 11 8 6 | 19
cue 21 24 15 | 55
Taken in seines by Mr. Seufert: |
HOD TAC h 69 Oltee imei scicie mete 136 112 149 135 285
Sept. 27, 1896 .--..-.-.-...._.. 157 146 186 160 343
Sept. 28, 1896. .--....--....... 117 ‘| 97 123 114 240
Sept. 29, 1896 2 102 73 115 83 217
Oct. 2, 1896 .. 137 123 166 145 303
Oct. 5, 1896 -- 149 105 144 123 293
Oct. 7, 1896 - - 96 | 83 127 119 223
Oct. 10, 1896 - 68 63 86 78 154
Oct. 13, 1896... 49 41 106 91 155
1,011 843 | 1,202 1, 048 2, 213

Silver salmon first appear in the southern end of Puget Sound abont the Ist of September, and the
run usually lasts until the first or middle of November. A few individuals are taken as late as
December, after which few are seen in the sound, but a good many are caught in Duwamish River.
23. Oncorhynchus nerka (Walbaum). Blueback Salmon; ‘‘ Redfish”; ‘ Sockeye.”

The investigations made in Idaho in 1894 and 1895 resulted in settling some of the disputed
questions concerning the redfish, but left others still in doubt. The details of the observations made
in those years will be found in the reports already published.* A summary of the conclusions reached
regarding the life-history of the redfish is given on page 16 of this article.

Although no satisfactory evidence was obtained as to the occurrence of the small form anywhere
in the stream below the lakes, this was not considered proof that it does not come up from the sea.
So close is its resemblance to the native trout, except at spawning time, that it would probably not
attract the attention of anyone, even if seen below the lakes.

In order to obtain more definite results in the study of this problem, the redfish lakes of Idaho
were again visited in 1896. Camp wus established July 11, on Alturas Lake Creek at the crossing of
the trail a short distance below Perkins Lake. Gill nets were set in the outlet the same day and were
kept set until September 25. These nets were examined from day to day during the entire period of
seventy six days and not asingle redfish was caught in them. The water was so high aud the current
so swift during the first ten days the nets were set that it was impossible to place them so as to wholly
obstruct the stream. It is therefore possible that fish may have ascended the stream and passed the nets

*A Preliminary Report upon Salmon Investigations in Idaho in 1894, by Barton W. Evermann,
Bull. U.S. F. C. 1895, 253-284.

A Report upon Salmon Investigations in the Headwaters of the Columbia River, in the State of
Idaho, in 1895; together with Notes upon the Fishes observed in that State in 1894 and 1895, by Barton
W. Evermann, Bull. U.S. F, C. 1896, 149-202, plates 67-72.

80 BULLETIN OF THE UNITED STATES FISH COMMISSION.

during that time. This, however, is improbable. Our camp was situated upon the bank of the stream,
the water was at all times very clear, and we were able to keep close watch for fish. If any redfish
had come up during those ten days it is more than probable they would have been seen by some of us.
After July 22 these nets were examined usually about three times each week, and though other fishes,
such as yellowbellies, Dolly Varden trout, suckers, and whitefish, were caught by them, no redfish,
either large or small, were taken. But small redfish were in the lake at least as early as July 16, and
they began entering the inlet August 3. Between August 3 and August 28 the number in the inlet
increased rapidly until the latter date, when 1,569 were counted.

The first small redfish seen was caught on a hook in Alturas Lake, just off the inlet, July 16, by
Mr. Maddren, while fishing for Dolly Varden trout. The hook was baited with salmon spawn. This
fish was a male 114 inches in total length, and the stomach contained a small amount of insect larve.
Of the many examples, taken by a gill net in Alturas Inlet on August 6, 9 were examined, only one of
which showed any trace of food in the stomach.

Twelve specimens, caught with grab-hook in Wallowa Lake about the 1st of September by Mr.
J.J. Stanley, were all quite fat,and food was found in the stomachs of all but three. This food
consisted chiefly of small crustaceans. a few insect larvie, and some gelatinous alga, probably a Nostoc.

Among the fishes from Lake Washington are 5 redfish, 3 of which are 7 to 8 inches long, the other
2 about 4 inches each. These were all collected June 15 by.Mr. Alexander. Each contains more or
less food in the stomach. Two other small redfish were obtained, which had been taken on the fly by
Mr. E. L. Kellogg while fishing in Lake Sammamish about May 15.

Whether this fish is anadromous or not is an exceedingly difficult matter to determine. If it comes
up from the sea it reaches the lakes much earlier in the summer or spring than has hitherto been sup-
posed. The fact that it feeds while in the lakes is now fully established, and it apparently continues
to feed almost or quite to the time when it runs into the inlets for spawning. ‘The one with food in
its stomach, taken in Alturas Inlet August 6, had just reached the inlet on that night. The specimens
from Wallowa Lake were caught in the upper end of the lake near the inlet, which they doubtless
would have ascended in a few days.

A consideration of small redfish from different localities, as to their size, proves interesting and
suggestive. Those from Alturas Lake are larger and much more uniform in size than those from
other places. Those from Washington, Stuart, and Nicola lakes are somewhat smaller, while those
from Wallowa Lake are much smaller. These differences in size are, in some cases, doubtless due to
differences in age, the specimens having been taken earlier at some of the lakes than at others, but
the marked difference between the Alturas and Wallowa specimens can not be accounted for on this
basis. It is doubtful if any of the Wallowa individuals would have reached even the minimum size of
those taken at Alturas Lake. This fact is brought out in the following table. In the first column are
given lengths in inches, the length being measured to the tip of the candal fin; in the othvr columns
are given the number of fish of each length from the respective lakes named at the head of the columns.

Table showing comparative sizes of specimens of the small redfish from different lakes.
Washing-
ton.

Chilowey- | Total

Length 3 PL:
= Stewart. Nicola. ate ATA,

ininches.| 4!turas. | Wallowa. |

a
OCRAOSCUUN RK WONRR EL

SALMON INVESTIGATIONS IN THE COLUMBIA RIVER BASIN IN 1896. 81

Table of comparative measurements of small redfish from Alturas and Wallowa lakes.

Wallowa Lake.

: i

No. pene Sex. Head. | Depth. | Eye. Snout. a Seales.
102 9.38 | Male.... 4.00 4.00 5. 00 4.20 1.90 i30
103 8.50 |..-do ...- 3.90 4.00 5. 00 3. 80 2. 00 130
104 (E16) sancti Fees 4.00 4.00 4. 67 4.00 1. 85 127
105 TeO00F |==- dose. 2 4.20 4.00 4.50 4.60 2.00 128
106 Shion does 4.00 4.00 4.75 4. 00 2.00 126
107 Biloee|aa Ole ee 4.00 4.00 4. 67 4.50 2.00 127
108 BiS8h Ee cdom ss 4.00 3.75 4.83 4.00 2. 00 120
109 OLO0Rs | Eaedareans 4.00 4.00 4.75 4.50 2. 00 126
110 8.63 |..-do.... 4.00 4.00 4.50 4.16 2.00 128
111 Bogor |. domes 3.83 4.00 4.80 4.00 2.00 122
112 i638 |e--doue-— 4.00 4,10 5. 00 4.00 2.00 126
113 6.38 |..-do-.... 4,40 4.50 4.25 5. 25 2.10 125
155 6525) |o--dovee > 4.40 4.50 4.33 5. 25 2.13 124

Alturas Lake.
No. aes ed Sex. Head. | Depth. Eye Snout. a Seales.
|

786 11.50 | Male.... 4.50 4.33 4.25 4.50 2.00

788 (27,00) ||-e2d0-=-— 3. 80 4.00 5. 00 3.50 1.80

807 12:°50). |= dos 3.75 4.00 5.25 3.33 1.75

808 11550") |Seedbiee= 3. 80 4.20 5.00 3. 33 1.80

808 12.00 | Female - 4.13 4.50 4.50 4.00 2.00

811 12.25 | Male..-. 4.20 4.40 5. 00 3. 67 1.83

810 12) ON) Eee Ges 3. 80 4.20 5. 20 3.50 1.80

783 12250 ier Ouse 3.75 4.33 5. 25 3.25 1.80

813 11.50 | Female - CAPER sees 5. 00 4.00 2.00

Specimen No. 155 was found dead on the shore at head of lake; specimens Nos. 113
and 155 with undeveloped sexual organs.

24. Salvelinus malma (Walbaum). Dolly Varden Trout; Western Charr; Bull Trout.

Obtained in Alturas, Pend d’Oreille, and Wallowa lakes, where it is abundant and affords good
sport for the angler, At Alturas Lake during July a great many were caught about the mouth of the
inlet and many large examples were taken in Lake Pend d’Oreille. The charr is a voracious fish and
preys largely upon other species of fishes. In the stomachs of some of those examined at Lake Pend
(’Oreille were numerous specimens of Leuciscus balteatus, Ptychocheilus oregonensis, Mylocheilus caurinus,
and Cottus. All of these except M. caurinus were found in the stomach of one fish. Some specimens
from ihis lake were infested about the axils by a small crustacean of the genus Lernwopoda.

Measurements of a number of specimens of charr are given in the following table:

Comparative measurements of charr from the Columbia River Basin.

Length in .
inches. | Head. | Depth. Eye. Snout. | Maxillary.
|
a@12.50 | 3. 80 4.80 6.40 4.00 2.00
ai8.50 | 3.40 6. 00 7. 50- 3.00 1. 60
612.25 | 3.83 5. 00 6. 60 3. 60 1.85
6, 50 | 4. 20 5. 00 5.25 4. 00 1.85
c5. 50 4.00 4.75 5. 00 4. 00 2. 00
d11. 50 4.00 5. 50 7.00 3. 83 2.00
d@10, 50 4.00 6. 00 6.00 3. 83 2. 00
d10.50 | 3. 83 6. 00 6. 00 3. 83 2.00
d11. 00 | 3.75 6. 00 6.00 4.00 2.00
a¥From Alturas Lake, Idaho. c From Wallowa Lake, Oregon.

b¥From Little White Salmon River, Washington. d¥ rom Lake Pend d Oreille, Idaho.

25. Salmo mykiss clarkii Richardson. American Cut-throat Trout.

Numerous specimens obtained. Particularly abundant in Lake Washington, Siuslaw River, and the
lakes near its mouth, Wallowa and Alturas lakes, Upper Klamath Lake, and in Des Chutes River. A
comparison of many specimens indicates that it will be necessary to recognize more species or
varieties of Salmo in the northwestern portion of the United States than have hitherto been admitted;
but it seems best to hold the matter open for additional data and consideration.

F. C. B. 1897—6

82 BULLETIN OF THE UNITED STATES FISH COMMISSION.

26. Salmo gairdneri Richardson. Steelhead Trout; ‘Salmon Trout.”

Seen in large numbers at Astoria and The Dalles. Not common in the Siuslaw, and only 2 speci-
mens obtained in Upper Klamath Lake. While at The Dalles during the last week in September
and the first half of October Mr. Alexander examined 4,179 steelheads, of these, 1,531 were males and
2,648 females; 476 males, and 900 females were well developed, and would probably have spawned in
four to six weeks. The remaining 2,803, he thinks, would not have spawned until some time in the
spring.

Steelhead salmon taken by Mr. I. H. Taffe, at Celilo, September 18 to 22, 1896, and examined by Mr. Alexander.

Male. Female.

How caught.
Date. Tw | Total. rk eS

Well de- Well de- In By
Number. veloped. | Number. veloped. 4 | wheel. spear.
SeptnlssiggG.se-eee eee ee 111 10 124 15 235 160 | 15
Sept. 19, 1896... 130 28 110 22 240 180 | 60
Sept. 21, i896...- oa. 121 29 366 98 487 192 295
Sapesion Tagg ee scereeeenee 159 59 391 164 550 350 | 200
521 126 991 299 1,512 882 | 630

Steelhead salmon taken in seines by Mr. F. A. Seufert, at Celilo, September 25 to October 18, 1896.

Male. Female.

Date. | Total
. | Well de- 4. Well de- :

Number. veloped. | Number. veloped.
Sept. 25, 1896...........seeeeee mA nSos6c 129 22 296 103 425
Sept. 26, 1896- st 148 28 281 112 429
Sept. 27, 1896. a4 218 91 328 140 546
Sept. 28, 1896.-----.-- sci52 97 44 140 48 237
ITY, VE eS neo an Ro Saeonsse ee poscscd 176 77 234 72 410
Oi ts Id 1GHGSs SoS enced cases isos ccesisstscd 49 | 12 50 14 99
(Ore Gitlin essa tessssconsr ss hasetesdsastete 116 49 181 69 297
(Ore y IRs 555 ees sees baecbssads Ses ssessssa7 46 12 81 21 127
Oot. 13, 1896) cccscescscvcserrsu=seervessencesss--- 31 15 66 31 97
1, 010 350 1, 657 601 2, 667

Mr. Alexander states that not many steelhead are seen about Seattle until the latter part of Novem-
ber or early in December, or about two months after they begin running up the rivers at the northern
endof the sound. During the early part of January, 1897, a good many were taken near Seattle.
They were considered equal in quality to those taken in other parts of the sound. Their eggs were in
various stages of development; a few fish were spent, but the majority were well advanced and would
have spawned in a short time.

27. Hypomesus pretiosus (Girard).

Nineteen specimens from Siuslaw River at Florence, where it is abundant about the cannery wharf
during the canning season.

28. Gasterosteus cataphractus (Pallas). Alaska Stickleback.

Found in Siuslaw River at Florence and at mouth of the outlet of Tsilteoos Lake. Only a few
specimens were obtained.

29. Gasterosteus williamsoni microcephalus (Girard). California Stickleback.

Many specimens from Tsiltcoos Lake, in which it is quite abundant. A few specimens were
obtained in Lake Washington by Mr. Alexander, where it was also obtained in 1892.

30. Siphostoma griseolineatum (Ayres). Pipejish.

One specimen, 10% inches in length, from the Siuslaw River near Point Terrace. It was caught in
a 7-inch-mesh salmon gill net.

31. Ammodytes personatus Girard. Sand Lance.

One specimen, 3 inches in length, from the Siuslaw River at Florence. Dorsal 60; anal 26.
32. Cymatogaster aggregatus Gibbons,

A Cottus which we with some hesitation refer to this species is very abundant in Siuslaw River
about tle cannery wharf during the canning season.

SALMON INVESTIGATIONS IN THE COLUMBIA RIVER BASIN IN 1896 83

33 Cottus asper Richardson. Prickly Bullhead.
Nine small specimens collected in Lake Washington, June 15. Head nearly or quite naked, but
entire body except belly uniformly covered with small prickles.

34. Cottus gulosus (Girard)

A Cottus which we with some hesitation refer to this species is very abundant in fresh and
brackish waters in Siuslaw River and in the lakes south of Florence, and easily caught on a trawl, or
with hook and line. Mapleton, Oregon, Siuslaw River, 25specimens; Acme, Oregon, South Slough, 4;
Whoahink Lake, 3; Tsiltcoos Lake, 20; Tahkenitch Lake,9; Lake Washington, 1; Lake Sammamish, 2,

35. Cottus punctulatus (Gill).
One specimen, 2.75 inches long, obtained from the stomach of a bull trout, Alturas Lake, July 16,
1896. Five from Alturas Lake, July 30, 1896.

36. Cottus princeps Gilbert.
Cottus princeps Gilbert, Bull. U. 8. F. C. 1897, 12, with figure.

Eighteen specimens from Pelican Bay, Upper Klamath Lake. Head 34; depth 5; D. vir or VIII,
20 or 21; A. 16 to 18; eye 5; snout 4; interorbital width 53. Ventrals 1, 4; lateral line scarcely
complete; anterior portion of body covered with prickles; posterior part smooth below. Head rather
long, pointed; maxillaries reaching to front of pupil. Color same as in C. leiopomus, which it closely
resembles, but from which it differs in the more numerous dorsal and anal rays, the shorter snout,
somewhat broader interorbital, and in having the body covered with prickles.

37. Cottus klamathensis Gilbert.
Cottus klamathensis Gilbert, Bull. U.S. F. C. 1897, 10, with figure.

Head 34; depth 44; eye 33; snout 34; D. yu, 19; A. 13 or 14; V. 1, 4; interorbital width 5;
maxillary reaching front of pupil. No teeth on palatines; a single straight preopercular spine, appar-
ently disappearing in older examples; nostril in distinct tubes; body short, stout; head moderately
broad, wedge-shaped anteriorly; caudal peduncle short, compressed, its greatest width 2 in eye, least
depth 1-in eye; gill membranes joined to the isthmus, widely separated; body smooth. Color rather
light; body with about 10 obscure irregular vertical blotches, but everywhere covered with minuet
brown spots, a V-shaped bar at base of caudal; spinous dorsal with a dark blotch on its posterior
portion; soft dorsal profusely covered with fine black specks; caudal faintly barred with dark and
lighter. A single small example from the lower end of Upper Klamath Lake.

Fic. 6.—UVranidea tenuis Fvermann & Meek, new species.

38. Uranidea tenuis, new species. ‘

Type No. 48229, U. S. N. M.; cotypes No. 5705, L. S. Jr. Univ. Mus., and No. 434, U.S. F.C. Type
locality: Lower end of Upper Klamath Lake, near Klamath Falls, Oregon, November 10, 1896.
Collectors Meek and Alexander.

Length of type, 3 inches, or 3g inches including caudal. Apparently allied to Uranidea marginata.

Head 34; depth 7; eye 44; snout 4; V. 1,3; D. vi-t, 17; A.15. Vomer with teeth; palatines tooth-

less. Head long, contracted from eyes forward; snout rather long; body much compressed, very
slender; greatest width of caudal peduncle 24 in eye; least depth of same slightly greater than eye;
preopercular spine well developed, broad, rather sharp, partly covered by skin, curved upward; below
this two other spines, the anterior one blunt, the other sharp, directed toward lower base of pectoral ;
post-temporal spine well developed. Body smooth, wholly without prickles or scales; lateral line
complete except on caudal peduncle,

84 BULLETIN OF THE UNITED STATES FISH COMMISSION.

Color, dark above and on upper three-fourths of sides, pale below; dorsal fins barred with series of
dark dots or blotches; caudalsimilarly marked but rather darker; anal light with a few dark blotches;
pectorals same ascaudal; ventrals plain; underside of head profusely covered with small, round black
specks; muciferous pores on head well developed.

Besides the specimen described above we have two others of the same slender style from Pelican
Bay, and many others from Pelican Bay, the majority of which are much smaller and less slender.
The slender ones are apparently entirely smooth; a ripe female 3 inches long has a few prickles on
anterior part of body, while all the smaller ones are pretty well covered with small prickles.

39. Leptocottus armatus Girard

Very abundant in Siuslaw River below Acme and easily canght on a trawl or with hook and line.
40. Platichthys stellatus Pallas. California Flounder.

Common about the mouths of Siuslaw River and Ten Mile Creek; occasionally taken during the
summer as far up the river as Mapleton. Of 11 specimens from Siuslaw River and Ten Mile Creek, 6
are dextral and 5 are sinistral.

LIBRARY OF CONGRESS

wi