NUTRITION AND GROWTH 



611 



were fed on a basal ration consisting of protein-free milk (containing 

 the inorganic salts, the sugars, traces of protein, and unknown substances 

 having an important influence on growth vitamines), to which were 

 added more carbohydrate, purified fat, and the protein whose influence 

 on growth it was desired to study. The same diet was fed at regular in- 

 tervals to a given batch of rats, and the weight of each rat was period- 

 ically taken, the observation being prolonged until the animals grew to 

 maturity and produced young, and these again grew to maturity, repro- 

 duced, and so on. By plotting the results in curves, with the time peri- 

 ods along the abscissae and the average weight of the rats of each batch 

 along the ordinates, the extent of the influence of a given diet on the 

 curve of growth was obtained. A normal curve of growth is shown in 



Days 



Each division - 80 days 



Days 



Each division -20 days 





Fig. 184. Curves of growth of rats on basal rations plus the proteins indicated. In curve 

 III the effect of the addition of zein to an inadequate allowance of the perfect protein, lactalbumin, 

 is shown; and in IV the effect of the addition of cystine to a deficient casein allowance. (From 

 Lafayette B. Mendel and T. B. Osborne.) . 



No. 1 of Fig. 183. It was obtained from results secured by adding liberal 

 amounts of casein to the basal diet. Similar curves were obtained with 

 lactalbumin of milk and ovalbumin and ovovitellin of egg. Perhaps the 

 most interesting substances capable of producing the normal curve of 

 growth are certain of the proteins that T. B. Osborne has succeeded in 

 separating in crystalline form from vegetable foodstuffs. These are 

 edestin (hempseed), globulin (squash seed), excelsin (Brazil nut), glu- 

 telin (maize), globulin (cottonseed), glutein (wheat), glycinin (soy bean), 

 cannabin (hempseed). 



That growth proceeds normally with any one of these proteins when it 

 is fed abundantly does not, however, necessarily indicate that each con- 



