38 American Fisheries Society 



The glass wool proved particularly ideal as it served as a filtering 

 medium, and could be ground to a thin powder, helping to reduce 

 the feces to a state of extremely fine sub-division and uniform 

 distribution. 



These powders, in which the glass represented many times the 

 bulk of the feces, were kept in weighing bottles and were easily 

 sampled for analysis. Nitrogen and fat determinations were 

 made on weighed portions in the usual manner. It may also be 

 mentioned that blank experiments were performed, and the 

 analytical data furnished corrections for the dissolved excreta and 

 feces which were found in experiments with trout. 



EXPERIMENTS IN FASTING. 



As a preliminary to the feeding experiments, a number of 

 experiments were performed on fasting trout. It has been found 

 that the alimentary tract usually frees itself of all excreta from 

 previous feeding in forty-eight hours. Minute quantities were 

 sometimes eliminated also during the next twenty-four hours, but 

 this was invariably negligible; the fasting, therefore, was generally 

 started forty-eight hours after the last meal. The first protracted 

 fast was performed with a trout weighing 102.9 grams (F-l) . In the 

 course of the four weeks of the experiment, no solid excreta were 

 eliminated. The water in the aquaria remained remarkably clean 

 for days, so that it could be changed at long intervals. The nitro- 

 gen eliminated in the water was determined for seven-day periods. 

 After twenty-eight days of fasting, the trout weighed 95.7 grams 

 or 7.2 grams less than at the start. The nitrogen eliminated during 

 the first week of fasting was 67.2 mg., but the quantity diminished 

 from week to week, only 45.1 mg. being eliminated during the last 

 week. The average elimination of nitrogen per day and kilogram 

 of weight was 81.3 mg. In the course of four weeks 233.5 mg. of 

 nitrogen was lost. It is evident, therefore, that about one-fifth 

 of the body loss was at the expense of the protein. 



The next two experiments were made with trout which had 

 been kept previously in the stock tank and fed freely. When 

 transferred to the experimental aquaria, these fishes vomited much 

 undigested food, and for a few days continued to eliminate large 

 amounts of feces. Trout F-2 in the first week of fasting lost 4.6 

 per cent of its weight, eliminating 141.9 mg. of nitrogen per day 



