as the author's experience, shows that the pigeon method for 

 studying 1 the vitamin B content of food, by the power of the 

 substance to prevent or cure polyneuritis, yields rather irregu- 

 lar results 15 ' 22 > 36 > 38 . The method of testing for vitamin B 

 by measuring the acceleration of yeast growth is open to the 

 question whether the culture medium is adequate in all other 

 respects and whether it is safe to assume that increased yeast 

 is attributable entirely to the vitamin 23 - 35 . The nutrition re- 

 quirements and rate of growth of 'the rat have been exten- 

 sively studied by Hopkins * 2j 36 , Osborne and Mendel 3 ~ 7 , 

 McCollum 8 - 11 and others. The rat method is therefore the 

 best standardized by previous work. Dr. Harriet Edge- 

 worth 37 , working in this laboratory, after a careful analysis 

 of the yeast and rat growth methods concluded : "The rat 

 growth method involves somewhat larger probable errors than 

 the yeast method, but can be interpreted in terms of B vitamin 

 with much greater certainty and is therefore the preferable 

 method." Skimmed milk in the form of dry powder was 

 chosen as the source of vitamin B, since it furnishes the vita- 

 min in a typical natural state, readily available and uniform. 

 It is easily and accurately manipulated in either large or small 

 quantities and is convenient for study in either the fluid or 

 dry state. 



EXPERIMENTAL PROCEDURE 



Selection and Care of Animals. The rats used were twenty- 

 eight or twenty-nine days old, this being the time of weaning 

 adopted in this laboratory, and weighed from thirty to sixty- 

 five grams when placed on the experimental diet. The experi- 

 mental animals all came from mothers on a diet of two-thirds 

 whole wheat, one-third whole milk and sodium chloride, two 

 per cent, of the weight of the wheat (diet 13) or this diet with 

 an addition of ten grams of raw lean beef per adult rat per 

 day (diet 13M). In each experiment, however, care was 

 taken, that equal numbers of comparison and control rats 

 came from mothers on the same diet. The lots were made up 

 in the following manner. If, for example, in one experiment 



