heated dry was fed in two ways: (1) mixed with the rest of 
the diet, replacing 25 per cent. of the starch in diet 94. This 
ration was always available and the rats ate ad libitum. (2) In 
another series of experiments the milk, which was heated dry, 
was reconstructed before feeding by making 25 grams up to a 
volume of 250 cc. with water. This was fed in doses of 8 cc. 
per rat per day. The advantage here is that all the rats re- 
ceived exactly the same amount of vitamin per day. Positive 
controls, animals receiving in the one case the diet containing 
25 per cent. of unheated milk and in the other case 8 cc. per 
day of the reconstructed unheated milk, were run in every case 
in comparison with the rats receiving the heated food. Nega- 
tive controls, receiving only the vitamin-free diet were run in 
almost all but not every experiment. This was done in order 
to make sure that the basal diet did not vary. 
THE ADEQUACY OF THE BASAL DIET 
In order to prove that the basal ration contained the 
optimum amounts of the various constituents, it was fed 
to a series of rats in comparison with diets in which 
each ingredient of the basal ration, the casein, butter 
fat and salt mixture, were in turn used in larger amounts. 
It was thought that perhaps another protein such as ex- 
tracted meat might prove to be a more adequate protein, 
or the meat might be more palatable and that therefore the 
rats would eat more of the basal ration. Consequently diet 94 
was also compared with a diet in which extracted meat residue 
replaced the casein. Commercial extracted meat residue was 
purified according to the method described for casein (page 7). 
TABLE I. 
The following diets were used in the comparison: 
Di€h varies eeeywies eens: 94 102 103 105 104 
% % Yo % % 
CaSO hye gin artisan paenig ieee 18 18 23 18 18* 
Bitter tat) scion. aiaveniesase ae 10 15 10 10 10 
Starch ..... err ee 68 63 63 67 68 
Salt mixture (O. & M.)*.... 0 4 4 4 b] 4 
“™* Meat residue. 
