Young rats of the Long-Evans strain were fed test diets of lard, lactose-free casein and salt mixture 185 for 3 weeks, after a preliminary observation period. The composition of the diets was 1.—87.5 per cent of the total caloric value from lard and 12.5 per cent from casein; 2.—75.0 lard and 25.0 casein; 3.—50 lard and 50 casein; 4.—25 lard and 75 casein; 5.—12.5 lard and 87.5 casein. The water intake was determined for each group on the stock diet, during the preliminary observation period and on the special diets. On diet 1, the water intake decreased 34.9 per cent; while it increased 14.3, 29.8, 22.0, and 34.2 per cent, on diets 2, 3, 4 and 5 respectively. On diets 1 and 2, the caloric intake, as compared with the stock diet, increased 11.3 and 19.5 per cent respectively. On diets 3, 4 and 5 the caloric intake decreased 15.9, 39.4, and 43.0 per cent respectively. The rats on diets 1, 2 and 3 gained steadily in weight, the mean total gain for the 21 days being 65.6, 80.7 and 42.5 gms. respectively. Those on diets 4 and 5 lost weight during the first week, but gained slowly during the second and third weeks. The mean total gain for the 21 days was 16.5 gms. for the rats on diet 4, and 13.0 gms. for those on diet 5. There was a significant correlation between caloric intake and gain in weight, except in the rats on diet 1, during the third week of the test period. The liver, in relation to the body weight, was significantly heavier in the rats on diets 4 and 5, than in those on the other diets. The livers of the rats on diets 1, 2 and 3 did not differ significantly from each other in relative weights. The percentage of liver glycogen was 1.64, 3.13, 2.73, 1.87, 2.44 on diets 1,2, 3, 4 and 5, respectively. The glycogen formed on the first 2 diets may be partly from the casein and partly from the glycerol of the lard, as shown by later experiments. The liver lipid content was 12.80, 5.62, 4.13, 3.05, and 2.94 per cent on diets 1, 2, 3, 4 and 5, respectively. In later experiments, adult discard rats were fasted for 48 hours, then fed test food, during the last 24 hours before killing. The glycogen percentage dropped to 0.137 in female rats and 0.513 in male rats, on fasting 48 hours. Glycogen was readily formed by feeding the stock diet, sucrose, casein, or glycerol. Only a small amount was formed from lard. The experiments in which sodium soap and fatty acids were used as test foods were unsatisfactory, as the rats could not be induced to eat any amount. In the one rat fed forced oleic acid, the glycogen content dropped below the fasting level.