588 REPORT OF COMMISSIONER OF FISH AND FISHERIES. [36] 
sack has nearly disappeared. At that time the lower jaw develops 
four strong, sharp canines, which curve backward and inward, forming 
stout hooks, two on the front of the jaw and two at the sides. It is evi- 
dent that these hooks are for the seizure and retention of a living prey. 
Such a provision would be useless for protozoa or rotifers, or any other 
animals as minute as these, and the fish itself is too small to eat any- 
thing as large as an amphipod crustacean. The whole apparatus is, 
however, well adapted for the capture of Entomostraca and minute 
insect larva, and it is very probably upon these that the little fish 
depends for its principal food.” 
The egg-sac of the fry in the warmer spring water disappeared some 
faster than with those in the hatchery, the fish developing in size to 
correspond; but their growth, however, was very slight. Buta few days 
after the absorption of the egg-sac the fish began dying, and by March 
1 were all dead. 
Those in the hatchery fared somewhat better. After the last lot were 
sent to Professor Forbes (March 14) about 40 or 50 remained, some of 
which lingered along until April 10, when the last one disappeared. The 
nourishinent derived from the scant animal life found in the water, sup- 
plemented by the meager sustenance afforded by the few particles of 
Gammarus eaten, although prolonging their lives several days after the 
egg-sac was gone, was quite insufficient to sustain life for any length of 
time. 
On March 18 about 200 young fish (hatched March 5), having their 
egg-sac nearly absorbed, were placed in a tank 24 feet deep by 14 in 
width and 4 in length, supplied with water from two one-inch spigots. 
These were offered Gammarus, as before; but as soon as the sac was 
entirely consumed they commenced dying quite rapidly, until only about 
20 remained on the 25th of April. These had grown a trifle in length, 
but were quite attenuated. A change of diet was then tried with bet- 
ter results to date (May 25) and prospectively. Liver and kidney, chop- 
ped into very fine particles, was substituted for the Gammari. The 
chopping process of preparing the food seems to be much better than 
pounding (the only way practicable with the Gammarus), as a minute 
subdivision of the mass is obtained without destroying the nutritious 
value of the particles by pounding out the soft tissue, leaving nothing 
but fiber. 
The fact that fragments of, Gammarus were found in the stomachs of 
the fry dissected by Professor Forbes would indicate clearly that their 
food must not be too impalpable before the gill-rakers are developed. 
This is corroborated by the conduct of the little fellows when fed the 
Gammarus “hash.” They would invariably attack the larger particles 
first—those quite too large to be swallowed; these would be heid by 
the mouth of the fish for a second or two, then spewed out and almost 
immediately seized again—an operation that was repeated until the 
particle had subsided to the bottom of the tank; then the smaller par- 
