520 Influence of Inanition on Metabolism. 



obviously increase markedly the percentage error, since any slight error in the 

 separation of feces would affect materially the coefficient of digestibility of 

 protein. 



Diet. Milk was the chief article of diet in this series of food experiments 

 in the calorimeter. In the nitrogen metabolism experiments a very much 

 more elaborate selection of foods was made, but they consisted in large part of 

 milk, cream, fruit, and vegetables, although occasionally meat was taken. 



Feces. In the discussion of feces after fasting it was pointed out that in 

 none of the calorimeter experiments here reported was there a sufficiently sharp 

 and accurate separation of feces made to enable any special quantity to be 

 designated as fasting feces. In only one instance was there any approximation 

 to such a separation and a subsequent consideration of the digestibility of the 

 food in the period following fast shows that in all probability the so-called 

 fasting feces were in reality a portion of the feces resulting from food preced- 

 ing the experiment. 



Separation of feces. It has already been stated that much difficulty was 

 experienced in obtaining sharp separations of the feces. In ordinary diges- 

 tion experiments where reasonably marked alterations in diet between the 

 experimental period and those preceding and succeeding it are concerned, the 

 technique of the separation of feces is one that is considerably more elaborate 

 than is commonly considered. In the case of the separation of feces imme- 

 diately following a fasting period the problem is even more complex, and 

 extreme difficulty has been experienced in nearly every instance in designating 

 the fecal mass that properly belonged to the period of experimenting with food. 

 As has been stated above, no typical fasting feces were isolated in these experi- 

 ments. Consequently, whatever errors were involved in the separation of the 

 feces between the preliminary food days and the first feces of the feeding 

 experiment proper increased considerably the liability of error in the proper 

 apportionment of feces to the experimental period. The marked alteration 

 in the character of the feces resulting from their long sojourn in the colon 

 was an added difficulty. 172 



Analyses of the feces. Ordinarily, in the analyses of feces, the determina- 

 tions of the nitrogen, carbon, organic hydrogen, water, ash, and heat of 

 combustion presented no unusual difficulties. The determination of fat, how- 

 ever, demands special comment. On the assumption that only small amounts 

 of soap exist in the feces, the amounts of fat have usually been determined by 

 extraction with ether. In the case of the feces from experiment No. 70 it 

 was found that the amount of fat was small, but when the heat of combustion 



1,2 In one instance 20.4 grams of feces were passed which contained but 50 per 

 cent of water. This is the lowest proportion of water ever found in feces in this 

 laboratory, and indeed we are not familiar with any records of fecal material that 

 contained so small a proportion of water as this. 



