NUTRITION. 381 



we could develop a suitable method by which the relative value of 

 different proteins, both for growth and maintenance, can be more 

 precisely determined. Experiments already in progress indicate that it 

 will be possible so to control the feeding that not only the absolute 

 amount of food and protein eaten, but also the amount in relation to 

 the body-weights of the animals, i. e., milligrams of food per gram 

 of body-weight, will be the same in all the comparable trials. 



It remains to be determined how far differences in the metabolism of 

 individual rats will affect the constancy of the results. Our experiments 

 indicate that this factor will have small effect, except in rare cases. 



The study of the efficiency of various fats in supplying certain 

 nutrition-promoting properties to the diets of isolated foodstuffs upon 

 which animals have been successfully grown has been continued. As 

 predicted in our report for 1914, beef fat obtained from some of the 

 abdominal tissues of the cow is not entirely devoid of the growth- 

 promoting properties found in butter fat and certain other natural 

 products. Our newer experiments have further shown that the failure 

 of lard to promote growth in the same manner as do other natural 

 fats {i. e., butter fat, egg-yolk fat, cod-liver oil) is not attributable to 

 deteriorating changes arising from heat or chemical agents in the com- 

 mercial manufacture of the product. Heating butter fat with steam 

 does not destroy its growth-promoting efficiency. Beef fat also renders 

 the inefficient diets used by us more suitable for producing growth 

 in rats than does lard. Wlien butter fat and beef fat are subjected to 

 fractional crystallization from alcohol, the growth-promoting factor 

 remains in the mother liquor or "oil" fractions. The fractions con- 

 taining fats with high melting-points are ineffective. The proportions 

 of added ''butter oil" and ''beef oil" fractions selected were usually 

 6 per cent of the entire food, obviously representing a much larger 

 addendum of the original fat. We have already shown that 6 per cent 

 of cod-liver oil also is satisfactory for growth. In comparing the 

 numerous records of growth on diets containing butter fat and beef 

 fat, respectively, we have gained the impression already referred to, 

 that butter fat is more effective in permitting growth than equivalent 

 quantities of the beef fat. Recoveries are less prompt and prolonged 

 growth is less satisfactory when the latter is used. In this connection 

 it may be observed that the yield of the liquid "oil" fraction from but- 

 ter fat is considerably^ larger than that from beef fat. The findings in 

 respect to the beef fat explain the fact, which we have observed, that 

 commercial oleomargarine also effects recovery in rats that have 

 declined on the lard diets. 



In evidence of the advantageous composition of the mixtures of 

 isolated foodstuffs which we have employed in recent years in our 

 studies on nutrition, we can now refer to records of albino rats which 



