DIGESTIBILITY OF THE GKAIN SOKGHUMS. 29 
grains — they may be considered to represent the digestibility of pro- 
tein determined directly rather than estimated. The carbohydrate 
portion of the diet was as well assimilated as in the ordinary mixed 
diet, indicating that the very incomplete digestibility of protein was 
not due to faulty experimental methods. The diet proved to be 
entirely satisfying, at least for the three-day period, for the subjects 
reported that they were in normal physical condition throughout the 
experiments. 
Comparing the experiments in which mush was eaten with those in 
which sorghum gingerbread formed the major portion of the diet, it 
was found that an average of 41 grams of protein per man per day 
was supplied in the former and 35 grams in the latter series of ex- 
periments. The explanation of such a small consumption of protein 
lies in the fact that the grain sorghums contain too little protein to 
make it feasible to obtain, say, 100 grams of protein daily on a diet of 
this nature. The amounts of energy supplied by the mush and bread 
diets were 3,000 and 2,850 calories, respectively, an energy value in 
agreement with the requirements of accepted dietary standards. 
SUMMARY AND CONCLUSIONS. 
In the experiments in which hard kafir bread was eaten with milk 
it was found that the protein of this cereal, as distinguished from that 
of the diet as a whole, was 58 per cent digested. Possibly owing to the 
large proportion of milk protein in the diet this value for the cereal 
protein alone may be too high. 
In the majority of the experiments the grains under consideration 
were eaten in the form of softer bread than that mentioned above and 
also in the form of mushes. Considering the grain sorghums studied 
the results show that on an average the protein of the softer dwarf 
kafir bread, as distinguished from the protein of the diet as a whole, 
was 51 per cent digested and that of the dwarf kafir mush 48 per cent. 
In the case of the protein of feterita the values were 51 per cent for 
the bread protein and 48 per cent for the mush protein. With the 
dwarf milo the values were 40 per cent for the protein of the bread 
and 34 per cent for that of the mush, and for the kaoliang 20 per cent 
for the protein of the bread and 4 per cent for that of the mush. 
Fat was present in these grains in very small and relatively unim- 
portant quantities and so no attempt was made to estimate the digesti- 
bility of this constituent as supplied by the cereals. 
In all cases it was found that the carbohydrates of the experimental 
rations and of the sorghums alone were very completely utilized. Con- 
sidering the grain only, as distinguished from the ration as a whole, 
the average value for the dwarf kafir (hard) bread was 98 per cent, 
for softer dwarf kafir bread 96 per cent, and for dwarf kafir mush 
96 per cent. In the case of carbohydrates supplied by feterita, the 
