effects of a varying oxygen supply. It may be that 
many other factors also are involved which depend on 
immeasurable quantities of dissolved materials. 
In the use of colored glasses placed between the 
eggs or fish and a source of light, it was observed that 
red light yielded results similar to those obtained in 
total darkness. Green and yellow lights seemed to be 
optimum for young salmon, while under the blue 
light there was the heaviest loss and the slowest rate 
of growth. These results are so much in accord with 
what one might suspect from the usually supposed 
effects of various light wave lengths that the tempta- 
tion would be to conclude that the shorter wave 
lengths are damaging to the fish. This is not borne 
out by the use of ultra-violet rays gave very diverse 
results, depending on wave length and time of ex- 
posure, ranging from favorable to adverse. If the fish 
were actually burned, death would result. 
When a sufficiently sensitive thermometer was used 
to observe the temperatures under each colored glass, 
it was found that the water was warmest under the 
red and coldest under the blue. It is well known that 
the speed of development of fish is governed by the 
temperature of the water, but very little was known 
about how small a difference of temperature was 
necessary to cause a noticeable difference in the rate 
of development. It appears from our experiments 
that probably less than o.i degree Centigrade will 
cause a noticeable effect. The observed temperature 
under the colored glasses varied more than this amount. 
Thus it is apparent that the differences in results were 
not causeo by light alone, if indeed light had any 
effect. The retarded development and heavier loss 
under blue glass can be accounted for by the lower 
temperature of the water. 
The oxygen contents of water vary with tempera- 
ture, and speed of development of the fish depends 
upon the amount of dissolved oxygen which the fish 
1901