62 FEEDING EXPERIMENTS WITH ISOLATED FOOD-SUBSTANCES. 



of rats on mixed food into the cages twice a week. It is an interesting 

 observation that when the rats kept on a mixture of isolated food- 

 substances were offered a choice between their own faeces and those 

 of rats on mixed diets, they invariably chose the faeces of the latter. 

 In many cases we have noticed a marked improvement in the nutri- 

 tive conditions of animals maintained on a single-protein dietary 

 when other rats were introduced into their cages for breeding pur- 

 poses. In view of the favorable influence exerted by feeding the 

 faeces of rats living on mixed food, it is quite likely that the presence of 

 the strangers in the cages furnished a suitable opportunity to obtain 

 "normal" faeces. This may explain the favorable results noted, in 

 contrast with the negative effects seen where several rats living on 

 the same single-protein diet have been maintained in the same cage. 



The extent of the influence exerted by what we have, in the 

 absence of a better explanation, assumed to be bacterial influences, is 

 illustrated in some of the appended charts, the periods at which the 

 faeces feeding was begun being indicated. The favorable effects have 

 not been confined to experiments w T ith one protein, but are mani- 

 fested with casein (see Charts XXXIX, XL, XLL and XLII), with 

 edestin (see Charts LXVI, LXVII, LXVIII. and LXIX), and with 

 gliadin (see Charts CI, CII, and CIII). Two failures may likewise be 

 recorded, viz, an ultimate one with casein (Chart XLI) and a com- 

 plete one with edestin (Chart LXXVII) as the protein component. 

 These were not due to incapacity of the animals to grow, since fur- 

 ther alteration of diet brought marked improvement. 



The influence of faeces feeding is especially striking in the case of 

 the gliadin tests, since without the addition of the faeces it has been 

 almost impossible to attain satisfactory nutritive condition with this 

 protein plus the special non-protein components of the food here 

 employed. It is instructive therefore to compare such failures (cf. 

 Period 2, Charts CXV and CXVI) with Charts CI and CIII, in which 

 faeces feeding was resorted to. 



In four of the experiments with edestin-food alluded to and re- 

 corded on Charts LXVI, LXVII, LXVIII, and LXIX, fresh faeces 

 were not actually introduced into the cages; but the improvement, 

 and even growth, in these young rats is coincident with the oppor- 

 tunity afforded to obtain "normal " faeces when other rats were daily 

 introduced into the cages for a few hours. 



In Chart CII is seen the result of an attempt to determine 

 whether the favorable influence of the faeces is actually of bacterial 

 nature. Faeces were fed as in the comparable gliadin experiments 

 (Charts CI and CIII) ; but they were previously sterilized by thrice 

 repeated heating in an atmosphere of steam. The decline of the 

 animal was not prevented to the same extent with sterilized as with 

 normal faeces. Further trials are necessary in this direction; and our 



