M or gulls. — Studies on Fish Nutrition 37 



retaining even minute particles floating in the water, but the 

 filtered water showed invariably a distinct turbidity. The tumbler 

 with the solid excreta retained in the glass wool was again dried 

 and weighed ; thus the weight of dry feces was gotten by difference. 



It may be mentioned that the sides and bottom of the aquaria 

 were thoroughly cleaned with a rubber tipped rod, and the wash 

 water added to the nitrate. This was acidified with five drops of 

 sulphuric acid and evaporated to a small bulk. As a rule, the 

 final volume was made up exactly to one liter. 



In view of the large quantities of water that it was necessary 

 to handle, the matter of evaporation presented certain technical 

 difficulties. Originally it was attempted to carry out the evap- 

 oration at a low temperature (50° C.) with the aid of a rapid 

 current of hot air. This method was very cumbersome, requiring 

 much time. Blank experiments have shown that there was no 

 particular advantage derived from the use of a low temperature. 

 A large steam bath was therefore installed, accommodating a 

 number of evaporating dishes (white enamel) of about twelve 

 liters capacity each. The steam bath was set in the apparatus 

 previously used for evaporating by means of the current of dry 

 air, so that the two methods could be combined at will. Very 

 large quantities of water could in this way be quickly condensed 

 to a small bulk. The condensed water was transferred to a vol- 

 umetric flask, the evaporating dish thoroughly rinsed with fresh 

 water, and the quantity brought up to a definite volume. This 

 condensed water was again filtered through asbestos to remove 

 such particles as may have gone through the glass wool. This 

 quantity was negligible, but the amount accumulated in the 

 course of an experiment was analyzed and added to the feces. 

 Aliquot portions of the perfectly clear water were used for analysis. 



In choosing fine glass wool as a means of separating the solid 

 excreta from the water, two important considerations were borne 

 in mind; owing to the small quantity of feces available, it was 

 very difficult to analyze it, and particularly to obtain a uniform 

 sample. It was urgent, therefore, first to mix the feces with some 

 other material to increase its bulk without interfering with the 

 analysis, and secondly, to make the grinding of the feces possible. 



The use of glass wool for this purpose suggested itself after a 

 number of different things have been tried with little or no success. 



