March 2, igoi.] 
FOREST AND STREAM. 
169 
reason of this slow growth in the upper stream, was due 
to the scarcity of food. Although the temperature of the 
water there is much lower than it is below or in the 
other streams, we cannot, knowing what we do of their 
growth in other cold streams, attribute their slow growth 
to the temperature of the water. 
Owing to this difference in growth in the same stream, 
the variation in the size of the individuals is great. Out 
of thirteen specimens taken from lower Hatchery Creek 
at the end of three and one-half months, the largest was 
.1-53. the smallest 2.26 inches, a difference of a Httle over 
1,25 inches; or, expressed in the amount gained by each 
since planting, the largest 2,18 inches, the smallest 0.91 
of an inch. The variation in Olema Creek, where no 
difference was found between the upper and lower 
stream, after the second month was nearly as great. 
From seventeen specimens taken at the mouth at the 
the stream immediately after hatching. Reaching matur- 
ity they would naturally follow the next run of salmon 
up the river. There are two runs of salmon up the 
Sacramento each year; the fall run starting up the river 
about the end of July and reaching the spawning beds 
at Battle Creek in October. The eggs spawned by this 
run hatch in January and February. Ten months after 
hatching, at which time the "grilse" would have matured, 
would be too late for them to catch the next fall run up 
the river, but they could catch the next spring run. which 
starts up the river in April, and reaches the spawning 
grounds in the McCloud River in June and July, the 
eggs hatching in October and November. The salmon 
hatching from this spring run get out of the river in 
time for the "grilse" to catch the fall run ten months 
later. 
If this is true the "grilse" running in the spring are 
utmost with food. Tomales Bay abounds in small crus- 
taceans and the young of the "silver smelt," and the 
young salmon would have no trouble in finding an abun- 
dance of food on reaching salt water. 
Scarcely any vegetable mattei" was found in their 
stomachs, and \vhat little was found no doubt was taken 
by accident. 
If these young salmon Hve almost exclusively on flying 
insects, the way to choose a good stream for planting 
would be to select one with plenty of trees, bushes and 
grass along its banks, for such vegeation is a harbor for 
insects. 
Enemies* 
The principal source of danger to young fish in a 
stream is from predaceous fishes, or even from older 
fish of their own kind. Of the four species of fish in 
PLATE n. — SALMON FRY. 
end of three months the largest was 3.42, the smallest 
2.58 inches. 
The important points learned in connection with the 
rate of growth of the young salmon were: First, that 
during the first four months their growth averaged 0.52 
of an inch per month; Second, that the few salmon re- 
maining in the stream after four months were stunted, 
and grew very slowly in comparison with those confined 
m Bear Valley Creek. 
An important fact to be noted in connection with the 
salmon confined in Bear Valley Creek is that of the two 
salmon taken in February, 1898, both were males; one 
with the generative organs fully developed. It is not 
uncommon to find young 4 to 4,^2 inch male salmon 
remaining in the streams with the testes fully developed, 
but it was never before found in a salmon whose age 
was known. This individual found in Bear Valley Creek 
was 5.5 inches in length, and was just twelve months 
old. It should also be noted that it was no smaller than 
PLATE in. — SALMON FRY. 
about sixteen months of age on reaching the spawning 
beds in June. Those arriving on the spawning beds in 
the fall are about twelve months of age. If this actually 
happens, the "grilse" of the spring run should average 
larger than those of the fall run. The age of these 
'"grilse" can be determined by better and more extended 
observations, and such knowledge would be a long step 
toward determining the important question of how old 
the full-grown salmon are when they run. 
Food. 
The young salmon, from the time they first begin to 
feed, eat .-^most anything they can get, but they shovi a 
preference for insects and insect larvae, and they pRcfe' 
to capture food that is floating on the surface of the 
water. An examination of the contents of the stomachs 
of the young salmon taken from Paper Mill Creek and 
its tributaries, from the time they were planted to the 
time they reached salt water, shows that their food was 
these Marin county streams — chub (Rutilus symmetricus) , 
stickleback {Gasterosteus cataphractus) , blub (Coitus 
gulosas). and trout (Salmo gairdneri) — the last two only 
are predaceous. 
After planting the young salmon, a number of these 
two species were caught daily and the contents of their 
stomachs carefully examined. Altogether only about 
twenty-five specimens of Cottus gulosas, of size large 
enough to prey upon the young salmon, were caught. 
Out of these twenty-five not one bad eaten a fish of any 
kind. Thirty or forty specimens of Salmo gairdneri were 
examined daily for three weeks after planting, and in not 
one instance had a salmon been eaten. The only fish 
eaten by them was Rutilus symmetricus, and no more 
than ten of these were found in about seven hundred 
examined. 
The other enemies to young fish observed were water 
snakes, kingfishers, herons, and divers. Although none 
of these were killed to find out what they were eating, 
the other male salmon taken at the same time. Evidently 
the development of the sexual organs did not retard its 
growth. 
Although these young male salmon are occasionally 
found with the generative organs prematurely developed, 
no females, to the best of my knowledge, have ever 
shown such development. This premature development 
peculiar to the males accounts for the undersized males, 
or "grilse," accompanying each run of salmon up the 
rivers, and it tends to prove that such "grilse" are not 
stunted individuals, but simply that they have matured 
at an earlier age and are younger than the salmon which 
they accompany. 
The Age of the "Grilse." 
The following is a mere speculation in regard to the 
age of the above mentioned "grilse": We have already 
seen that one out of the two males taken in Bear Valley 
Creek af the age of one year was sexually mature. In 
the headwaters of the Sacramento we have in November 
found among the A^oung salmon remaining in the streams 
several sexually mature males, which at that time were 
under ten months of age. These few examples which 
have come to my notice represent a very small per- 
centage of the number that • must mature within ten 
ippriths Pf 3 year \r\ ^J^e l^ge ^odjr of ^al^^on th^t leave 
almost exclusively flying insects which had dropped 
upon the surface of the water. 
These streams abound in caddice worms of three or 
four species, small periwinkles, the larvae of stone-flies, 
may-flies, and other insects; but they were all neglected 
for the flies, bees, beetles, caterpillars, etc., that fall into 
the stream. No caddice worms or periwinkles were 
found in their stomachs, and but very few of the larvse 
of aquatic insects. This same thing was observed in the 
younger trout; and although the larger trout eat large 
quantities of caddice worms, the insect larvae in the water 
do not furnish the amount of food to these fish that is 
popularly supposed. 
Although these salmon live almost exclusively on in- 
sects that drop upon the water, we have found in their 
stomachs pieces of leaves and buds, small feathers, shells 
of salmon eggs, and the helpless j'oung of their own 
species. (See notes on planting of salmon fry before 
yolk sacs were absorbed, at Sisson, November, 1897.") 
In the salmon taken in brackish water I found they 
had been eating small salt water crustaceans, and one 
three-inch salmon taken in salt water had eaten six of 
the younsr of the "silver smelt" (Atherinops aMnts) and 
one small leaf-hopper of the kind that is found among 
the pebbles along the shore of the bay. 
In atoost every casp the ptom^Gh? Yfere fiH^4 to 
it is safe to say that the loss of young salmon due to 
them was slight. 
Planting of Quinnat Salmon Fry in Marin Coonty Steams 
in J89S. 
In February, 1898, 2,000,000 quinnat salmon fry were 
planted in Paper Mill Creek and its tributaries by the 
United States Fish Commission. On account of limited 
hatchery space and lack of funds, the fry were all planted 
before the yolk sacs were absorbed. The fry hatched 
at the same time as those liberated in these streams last 
year, but they were planted about three weeks earlier. 
The height of the water in the streams at the time of 
plantmg and during the next month was lower in 1898 
than m 1897. Notwithstanding this, the salmon ran out 
earlier than the year before. In April there were not as 
many salmon left m the streams as in June of the year 
before. In June no salmon whatever could be found in 
Paper Mill Creek. Only a very few were found in Olema 
Creek, and they were not much more plentiful in 
Hatchery Creek— the same condition that existed two 
months later in 1897. The only reason I can give for 
this difference in time of running out is that they were 
planted at different ages. Apparently, planting before 
the^yolk sacs are absorbed causes them to run out earlier 
This liriiig§ an important quest|e!i|3 Does tfie holding 
