4 BULLETIN 470, U. S. DEPARTMENT OF AGRICULTURE. 
of the total. In the case of the other grain sorghums the values 
were, milo, 19 per cent; feterita, 15 per cent; and kaoliang, 5 per 
cent. That these different proportions of the four meals prepared in 
the same mill should be obtained under similar conditions is most 
likely due to the structure of the grain, for, although the break was 
made in the same way, the different grains did not fracture evenly 
along the same lines. The protein content of the meal, which de- 
pends partly upon the quantity of bran removed in the process of 
milling, was proportionately less in the kafir and milo meals than in 
the others, since more bran was removed from them on the sieve. 
COOKING TESTS AND THE EXPERIMENTAL DIETS. 
Before deciding upon the experimental diet a series of tests was 
undertaken to determine a good method of preparing the sorghum 
meals for eating and also a suitable combination of foods to serve as 
a basal ration in the digestion experiments. Results previously re- 
ported from the Office 1 of Home Economics indicated that dwarf 
kafir requires a more thorough cooking than is customary for Indian 
corn, and that it is often desirable to make a mush of the meal as a 
foundation for other dishes (yeast bread, gingerbread, etc.). In view 
of this experience, a mush was first made of one of the sorghum meals 
(kafir) by cooking 3 cups of meal, 3 cups of water, a little salt, and 
3 tablespoonfuls of lard for a half-hour in a double boiler. The lard 
was simply placed on top of the other ingredients, not mixed with 
them, for it was noted that when the lard melted it floated on top 
and prevented excessive evaporation. Baking powder was then 
beaten into the mush and the bread was baked, but because of the 
lack of gluten in the kafir meal difficulty was experienced in making 
such bread light. The addition of glutinous flours, eggs, milk, or 
other substances containing protein was not desirable, for it was recog- 
nized that the presence of protein from any other source than the 
sorghum grain would complicate the problem of determining the di- 
gestibility of the grain protein. It was found, however, that by the 
above method suitable bread resembling corn bread could be made, 
providing the batter was baked in a very thin layer. Owing to the 
loss of practically all the water originally present the bread was 
extremely dry, hard, and crusty and contained practically no crumb. 
It was used only in the first series of experiments with kafir meal, 
when this sorghum was eaten with a basal ration containing only 
milk, fruit (oranges), and a little sugar to make the diet palatable. 
The experience gained in these and earlier cooking tests made it 
evident that in order to cook the grain sorghums thoroughly a pre- 
liminary softening by boiling is desirable, or else it is necessary to 
have sufficient water present to soften the grain during the baking 
i U. S. Dept. Agr. Fanner's Bui. 559 (1913), pp. 6, 7. 
