488 D. W. CUTLER. 
1916, they both died in an apparently exhausted condition. The same 
applies to the scales 8. 8, 8. 10, being from fishes 8 and 10 of the scanty 
tank. It will be seen that the curves show the salient features of all 
those for that period taken from fish living in unheated water. They 
are characteristic, however, in showing the possession of very few 
sclerites, and in having the width of the sclerites slightly less than 
is usual, 
Scale curve H. 15 is the most interesting in showing the predominance 
of temperature over nutrition in regard to sclerite form. It is a curve of 
a scale from a fish living in the hot tank, No. 15. From January to 
May the animal was in fairly good condition and produced 10 broad 
sclerites : from May to October, however, its condition was bad, but 
in spite of this six sclerites were produced and instead of being narrow, 
as might have been expected if food were the cause of width variation, 
they retained the broad width of 1:3. 
In all cases the same effect is seen, that the lower nutrition, leading to 
a poor condition, results in the lessening of the number of sclerites pro- 
duced, but does not affect the breadth. 
Reference to the tables will show that where the condition of a fish is 
reported as poor, very thin, ete., the number of sclerites produced is 
small, but the maximum and minimum widths remain similar to those 
for fish in good condition. 
CONCLUSIONS. 
The conclusions which I draw from the result of these experiments 
on the scale growth of fish is, that the broad summer bands, which are 
caused by the sclerites formed during that period being wide, and the 
narrow winter bands, produced by narrow sclerites, are due to changes 
in the temperature of the water in which the animals are living. High 
temperatures, such as are found in the summer months, lead to the 
formation of broad sclerites, while the narrow ones are called forth by 
low winter temperatures. 
Owing to the temperature of the water varying from month to month, 
or even from week to week, the scale curves do not show a continuous 
rise and fall, but exhibit at certain places secondary elevations or depres- 
sions, which I have termed secondary maxima and minima. These, I 
believe, to be due entirely to the above-mentioned variations in the 
monthly temperatures. 
The amount of food which the fish consumes and its general condition 
does not afiect the production of summer and winter bands: the only 
effect which poor nutrition seems to have on the scales is a tendency for 
the production of few sclerites. High food consumption leads to a high 
sclerite formation. 
