58 FEEDING EXPERIMENTS WITH ISOLATED FOOD-SUBSTANCES. 



of proteins. Sodium chloride obviously has no potential neutraliz- 

 ing power. If the usefulness of the salts were associated solely with 

 their specific character as salts, the salts of sodium ought to be some- 

 what comparably efficient. 



The function of the inorganic salts is by no means exhausted, how- 

 ever, by the simple action of chemical equilibrium. It would lead us 

 too far afield in this place to discuss the problem in detail. Charts 

 XI, XII, and XIII, Part I, pp. 38-39) showing the marked differences 

 induced by alterations in the inorganic salts of the diet, the other 

 food components remaining unchanged, are highly suggestive. We 

 have since then made numerous attempts to improve upon the salt 

 mixture empirically selected and prepared somewhat in imitation of 

 the ash of milk. Rats were kept alive (while they steadily declined) 

 84 days on a food mixture which analysis showed to contain only 

 minimal, inevitable traces of ash (0.16 per cent, a considerable part of 

 which was phosphoric acid derived from the casein) . Chlorides were 

 entirely lacking, distilled water being furnished for drinking. In 

 view of this it is necessary to proceed with extreme caution in draw- 

 ing conclusions from observations extending over brief periods. We 

 shall refer to the subject again, it being sufficient here to emphasize 

 the subtle and specific value of the salts. The lack of knowledge in 

 this field has furnished an obstacle which we have only lately suc- 

 ceeded in overcoming in part. 



Even when all these varied conditions are taken into account, 

 there still remain, as we have pointed out before, extraneous inci- 

 dents and accidental factors apart from nutrition itself, which may 

 complicate or vitiate experiments like those projected. Disease, old 

 age, injury, may be mentioned in illustration. Failures to maintain 

 nutrition successfully under such extreme conditions do not neces- 

 sarily imply a deficiency or inadequacy of the dietary. Accordingly, 

 successful experiments must be given far greater weight than failures, 

 where so many possibilities of detrimental influences, aside from the 

 diet itself, are liable to arise over prolonged periods of observation. 

 Some of the uncertainties have been eliminated by the experience 

 previously gained. For example, the intercurrent diseases of our 

 animals have been almost entirely excluded by the use of rats raised 

 in the laboratory for this research. By the prompt elimination of 

 diseased animals, by scrupulous attention to the conditions of the 

 cages and feeding arrangements in other words, by painstaking 

 attention to hygienic factors we have succeeded in maintaining a 

 large number of animals in exceptionally good health, so that they 

 have become the more suitable to permit of accurate conclusions 

 regarding the effects of the diets studied. Furthermore, the age and 

 hereditaryfactors in our animals are nowknown to us, so that another 

 source of uncertainty has disappeared. 



