248 E. V. McCOLLUM 



course, always easy to calculate from the amount of food consumed by an 

 infant whether it is taking- sufficient food to support growth, or whether 

 it is on a maintenance or submaintenance diet so far as energy values are 

 concerned. 



Kosenstern urges, reasoning on the same basis, that a child of a weight 

 considerably below that which corresponds to its age, will need a higher 

 intake of calories per kilogram for an increase in weight due to symmetri- 

 cal growth than either an infant of the same weight but younger, or one 

 which is of the same age but heavier, i. e., normal in weight. Finkel- 

 stein(c) (cited by Aron) advises that a child be given the number of cal- 

 ories corresponding to its age, irrespective of its weight. Such a general- 

 ization presupposes, naturally, that inhibiting factors other than lack of 

 sufficient food are discovered and eliminated. 



Osborne and Mendel have shown that rats fed ad libitum on diets which 

 were satisfactorily constituted except that the amount of protein was so 

 small as not to permit of growth, or the protein was present in relative 

 abundance but was of such poor quality that only maintenance with- 

 out growth was possible, could be maintained at constant weight for 

 surprisingly long periods, in terms of the span of life of the species, 

 without their losing their capacity to grow. When the diet was corrected 

 the animals grew at a rate faster than the normal, but failed to reach the 

 maximum size which might have been reached under better conditions 

 of uninterrupted nutrition. It is not warranted, as has already been indi- 

 cated, to generalize from these observations, concerning the retention of 

 the capacity of an animal to grow after periods of suspension from any 

 cause. In the case of calcium at least, we have a dietary factor of which it 

 is of the utmost importance to maintain an uninterrupted satisfactory 

 ingestion, for a temporary suspension of growth due to lack of this ele- 

 ment leads to permanent deformity. 



It is possible, in the light of experience in experimental work with 

 the rat, to start with a diet consisting of degerminated cereal products 

 such as bolted wheat flour, degerminated corn meal, and polished rice, 

 the remainder being made up of peas, beans, potato and rolled oats or 

 other foods having similar dietary properties, and to so modify it in a 

 series of experiments that the span of life, fertility, development of the 

 sucklings, infant mortality and age at which signs of senility appear, can 

 be made almost anything we desire. The mixtures described, irrespective 

 of the proportions in which they are combined, will not support any growth 

 whatever in a young rat. The deficiencies have been shown to be a lack of 

 sufficient calcium, phosphorus, sodium, chlorin, fat-soluble A, and low bio- 

 logical value of the protein mixture. The importance of the several factors 

 named are in the order given. 



If we systematically enhance such a standard mixture with 

 respect to the factors named, but in limited degrees, in a series of 



