Morgulis. — Studies on Fish Nutrition 35 



through the water and was dissipated as a fine spray by the 

 aeration tube. The air escaping from the jar bubbled through the 

 standard acid contained in the cup. This precaution was taken in 

 order to avoid any possible loss of ammonia from the water by the 

 vigorous current of air. 



The contents of the cup were emptied every twenty-four hours 



N . 

 and titrated with r^jn s °dium hydroxide. As no changes in the 



quantity of acid have been found in the course of many trials, 

 this practice was discontinued as unessential to the accuracy of 

 the experimentation. 



The aquaria were submerged in a trough of galvanized iron 

 through which a constant circulation of water of fairly uniform 

 temperature was maintained. The temperature of the aquaria in 

 which the trout sojourned was thus regulated. 



The water used in these experiments was exceptionally pure, 

 coming directly from the excellent filters installed in the New 

 York Aquarium. No sediment was formed by this water even 

 upon standing several weeks. 



The trout which were experimented on were weighed at the 

 beginning and close of each period. Different methods were 

 tested for obtaining the accurate weight of the fish, and finally 

 the following method was adopted: The trout was picked up 

 with a small, fine net, the adhering water shaken off and allowed 

 to drain for about half a minute. The fish was then transferred 

 cautiously to an aluminum can partly filled with water, the 

 weight of which was accurately determined. With practice it was 

 possible to perform this operation without splashing a drop of 

 water. The can was closed and again weighed with the trout, 

 whose weight was thus gotten by difference. Though the method 

 is not free from certain defects, it had two important advantages 

 over every other method tried, in that, in the first place, no 

 injury was done to the trout; and secondly, with the animal 

 securely in the can, the weighing could be done leisurely. Of 

 course it is assumed that the amount of moisture adhering to the 

 animal has been the same at each weighing. Though this assump- 

 tion is arbitrary, it has been found, by weighing the animals sev- 

 eral times in succession, that the extreme differences do not vary 



