74 



THE AMEBIC AN NATURALIST [Vol. LV 



from different sources. In tliis connection, it is believed that the 

 procedure to be described later by the writer jjives more direct 

 evidence and hence is more free from criticism. 



In order to eliminate the question of variability of individual 

 animals from the point of view of efficiency as ener^ trans- 

 formers, it appeared to the writer that a better procedure than 

 that used by Hopkins would be to feed a number of animals a 

 basal diet plus such an amount of vitamine-containing material 

 as to keep the animal in weight equilibrium over a period of 

 time. The amount of vitamine would vary with the individual, 

 and would need to be determined in each case. When the animals 

 had been maintained in weight equilibrium over a number of 

 days and the food consumption not^d, they could then be fed the 

 same daily ration of a diet containing the same number of 

 calories and having the same composition with the exception of 

 a larger vitamine content and the weight noted. As the greater 

 vitamine content would stimulate the appetite (directly or in- 

 directly) , there would be no difficulty about the animals eating 

 the same amount of food as they had eaten in the first stage of 

 the experiment and they would receive the same number of 

 calories and a food with the same gross composition within very 

 narrow limits. 



Due to lack of time available for this work, it was impossible, 

 in the preliminary experiment to be described, to bring each 

 animal as near weight equilibrium as was wished, and as it is 

 felt verv^ certain can be done. The average, however, for twelve 

 animals is close and if the results are considered from the statisti- 

 cal point of view, they give further very convincing evidence that 

 the vitamines increases the efficiency of the body in the utiliza- 

 tion of the food. The error due to temperature variation it is 

 believed is not large, but this, of couree. should have been elimi- 

 nated. There appears no veiy easy way of eliminating the error 

 due to greater activity of the animals in the second stage of the 

 experiment. Tliis error may be considerable for there is no 



the liiiiitc.l .liet with greater vitamine content. The direction of 

 this error makes the results all the more convincing. 



Experimental 



Preparation of Diets. — The diets were prepared at the be- 

 ginning of the experiment from the same stock of material and 



