486 D. W. CUTLER. 
the second, though normal as regards the width of its sclerites, is remark- 
ably deficient in them. 
The fish to which scale No. 12 belonged was in very good condition 
and young: this was not the case with the fish from which No. 15 was 
taken. I attribute these peculiarities to nutriment, but will deal especially 
with this aspect of the problem in a later part of the paper. 
The growth of the fish in the cold tank was, on the whole, very good. 
Table IV, which deals with the details of the fishes experimented upon, 
seems to show that the increase in length was slightly greater than that 
of the animals in the hot tank. There are, however, possibilities of error, 
such as faulty measurements and the small number of fish experimented 
upon, which make it impossible to draw any definite conclusions. 
The figures of the maxima and minima for the sclerites for each period 
arc, however, striking when compared with those of the corresponding 
periods for the hot tank animals. The figures are also lower than those 
for the control tank, though the difference is not very marked. This is, 
however, to be expected when it is remembered that it was only possible 
to keep the temperature of the cold tank a few degrees below the normal. 
The scale curves, Fig. 7 (p. 479), are very interesting in showing the 
marked difference between them and those of the hot tank for the 
period of January to May. In all cases, and this was found in 
scales not figured, the curve never rises above unity, while in the hot 
tank it never falls below that width. 
For the period January to October, the curves have a tendency to 
rise ; there is an increased width of sclerites for the months July to 
September, but the greatest width does not approach that of the hot tank 
fish for the same months. 
Compared with the control scale curves, there is similarity, but we find 
that the widths are more uniformly low. 
When considered in relation to the temperature for each month the 
curves follow very closely the varying changes in the degree of heat, which 
the water possessed in which the animals lived. 
As in the case of the hot tank scale curves we note that these curves 
also follow one another very closely, only differing in minor degrees. 
The results of the experiments on the scales of fishes living in artificially 
heated and cooled water seems to indicate clearly that the temperature 
of the water has a very marked effect on the width of the sclerites pro- 
duced. It is well to remember in connexion with this that the feeding of 
the fish in the two tanks was exactly the same. 
ABUNDANT AND SCANTY TANKS. 
As I have already said in these two tanks the temperature of the water 
was not altered. The only difference in the way the fish were treated 
