Morgulis. — Studies on Fish Nutrition 47 



The amount of each constituent consumed was then determined 

 by the difference. 



In his study of the digestion of fish, Knauthe* made use of 

 this method exclusively. Apart from the disadvantage which it 

 shares with all indirect methods, the method is time-consuming 

 and presents so many unforeseen difficulties as to render the 

 results at times valueless. 



In the matter of recovering the fat, Knauthe's procedure, 

 though very simple, consisting merely in extracting an aliquot 

 portion of the water with ether, on closer examination proves of 

 little or no value. Knauthe has not performed blank experiments 

 to determine how closely the recovered amount corresponded with 

 that washed out by the water ; he therefore had no occasion to be 

 apprehensive about the acceptability of his results. 



As was pointed out in an earlier section, I considered the per- 

 formance of blank experiments as very essential, and have relied 

 on these in deciding whether or not the analytical procedure in use 

 was adequate. By following Knauthe's method, I found that the 

 results were so widely off the expected value as to make them 

 absolutely worthless. 



It will be noted further that in all the experiments, the animals 

 were allowed to remain in the feed jars one-half hour, after which 

 they were removed to fresh water. Knauthe on the other hand, 

 allowed his animals to remain in the aquaria containing food for a 

 very long time, and separated mechanically the feces from the 

 solid food particles. It need hardly be pointed out that such a 

 procedure is altogether too crude to warrant great confidence in 

 the significance of experimental results. The bacterial growth in 

 suspensions of nutrient material would be sufficient to vitiate the 

 results. In a number of blank experiments a weighed quantity 

 of liver pulp was put in the aquaria (without trout), left there a 

 half hour as usual, then the solid residue and the filtered water 

 were analyzed. In the case of the nitrogenous material and the 

 sugar, the recovery was complete within less than one per cent, 

 but in the matter of the fat, as determined by direct extraction of 

 the water with ether, such low values were usually gotten as to 

 render the data of no importance. The need of a different and 



*Zeitschrift f. Fischerei, Vol. 6, p. 139, 1898. 



