PROGRESS IN BIOLOGICAL INQUIRIES, 126 539 
as roughage and is entirely unsuited to feeding small fish. The dried 
shrimp used at Holden is a special grade, consisting of the abdomen 
only, which has been freed from the surrounding shell. After being 
soaked in water for several hours, it can be ground sufficiently fine 
to serve as food for fish 2 to 3 inches long and upwards. When 
used alone it is inadequate, as shown by a sharp increase in the mor- 
tality; but the addition of sheep liver quickly brought the mortality 
back to normal. A continuation of the experiment on the basis of 
>0 per cent shrimp and 50 per cent sheep liver resulted in practically 
no losses and an exceptionally rapid growth. 
The results with “clam” meal were particularly encouraging, 
although, owing to the small amount available, it was possible to 
try it on a small scale only. A lot of steelhead fingerlings fed beef 
heart with 25 per cent “clam ” meal over a period of 97 days showed 
a mortality less than one-half as great and a growth over one-fourth 
greater than the controls fed beef liver. The meal is made from the 
dried “ meats” of fresh-water mussels and is a by-product of the 
mussel-shell industry. Inasmuch as it is impossible to obtain a satis- 
factory grade of this meal in quantity. an attempt will be made to 
manufacture it at the Fairport (Iowa) biological station. 
A series of experiments to determine the effect of sunlight on 
young trout yielded very interesting results. Various lots of fish 
from the same source were placed outdoors in different compartments 
of the same trough, with some compartments covered so as to exclude 
all direct light, others were covered with ordinary window glass, 
which would absorb the ultra-violet rays, while some compartments 
were exposed entirely. All lots were fed alike, and the flow of spring 
water was abundant, so that on the hottest summer days there was no 
appreciable difference in temperature between head and foot com- 
partments of the same trough. The only difference in the various 
lots was the rather abundant algal growth in the exposed compart- 
ment, but it is not thought that this would react to the disadvantage 
of the fish. 
Lake trout in advanced fry and fingerling stages showed losses of 
18.4 per cent in the covered lot and 33.7 per cent in the exposed lot. 
A repetition of the experiment, using advanced fry and fingerlings 
of rainbow trout, gave even more striking results. In this case, over 
a period of about two months, the protected fish suffered a loss of 
only 1.2 per cent; in the glass-covered group. the loss was 3.5 per 
cent; while among the fish that were entirely uncovered there was a 
mortality of 38.1 per cent. No record of the weights of the lots was 
taken, but plainly the condition of the fish was correlated with the 
mortality, inasmuch as the groups exposed to sunlight, which suf- 
fered the greatest loss, showed poor condition and comparatively 
slow growth. 
On the face of the results it would appear that ultra-violet light 
is injurious to trout, at least in the younger stages, and that the 
common practice of protecting trout at the hatcheries from direct 
sunlight is based on sound reasoning. It should be pointed out, 
however, that the fish were held in shallow troughs and that the 
results may not be applicable to fish held in ponds or raceways. 
Further experiments on the influence of light are planned for the 
near future. 
