lana 
CHEMISTRY AND PHYSICS. 47 
presented to this Academy the evidence that individual stalks of sorghum vary 
greatly in their sugar-content, even when of the same variety and general de- 
velopment. That fact followed up by the Kansas Experiment Station and the 
United States Department of Agriculture resulted in the improvement of several 
varieties of sorghum, by selection of individual stalks by chemical analysis of the 
juice, to the extent of several per cents. That similar results may be obtained 
with corn seems not unreasonable. With corn, however, the desire is to increase 
the percentage of proteids in the grain. Corn is king, no doubt, now, but we 
desire to make him more worthy still of his regal position. Corn is rich in car- 
bohydrates and fat, but too deficient in proteids to make it an ideal feed for any 
purpose. Its best utilization requires that its excess of carbonaceous constityents 
be balanced by the addition of feeds rich in nitrogen; hence the dairyman adds 
wheat bran, gluten meal, oil-cake, or some other feed rich in proteids. For 
growth, for labor, and even for fattening, corn lacks nitrogen. The average per- 
centage in corn is about ten; in wheat, twelve; in bran, sixteen; and in oil-meal, 
thirty-three. Could we increase the proteids of corn by two or three per cent., 
it would raise the value of the annual crop millions of dollars. 
The compilation of analyses of corn published in Experiment Station Bulletin 
No. 11, office of experiment stations, shows that the average protein-content of 
dent corn is 11.5 per cent., calculated on the dry matter; the minimum is 8.2 per 
cent., and the maximum 13.8. 
The results about to be given were obtained in analyses incident to a joint ex- 
periment now being carried on by the farm, botanical and chemical depart- 
ments of the Experiment Station of the State Agricultural College, in which the 
object is to improve corn in its protein-content, by seed selection based on chem- 
ical analysis. Cross-fertilization between the best varieties, it is hoped, will 
result in the establishment of a better one than any now known, and careful 
selection should lead to relative fixity of type. 
As a preliminary, analyses were made of single ears from thirty-three varieties 
grown in this state, and collected as good ones. These showed a nitrogen-con- 
tent of from 1.56 per cent. to 2.26, corresponding to 9.75 and 14.12 per cent. of 
proteids, respectively, calculated on the dry substance. These would not be fair 
figures by which to judge the varieties, as but a single ear was used for the sam- 
ple, and it was found by another set of analyses that individual ears of the same 
variety varied to as great an extent. 
To study the variation in different ears of the same variety, one was chosen 
which was the result of a crossing of white and yellow corn, and another which 
had been grown on the same farm by the same man for thirty years, without ad- 
mixture of other seed. It was thought that the former would represent an un- 
stable variety, and the latter a fixed type, if such could be formed by ordinary 
selection of seed. Ten individual ears of each were analyzed. The cross-bred 
corn showed a variation from 1.35 per cent. of nitrogen to 2.22; the other varied 
from 1.53 to 2.24. These great differences, even in the case of the supposed fixed 
type, show what wonderful possibilities in seed-selection are open to us. If the 
average composition can be raised to that of the best ears, the grain will be equal 
to wheat in feeding value, if equally digestible. In sampling for these analyses 
a narrow belt of kernels was taken around the middle of the ear in each case. 
We went still further, and took the ear from the fixed type which had shown 
the highest percentage of nitrogen, and made analyses of individual kernels from 
it. These were taken from the middle of the ear, and were selected to as nearly 
as practicable the same weight, since kernels of differing size might easily be 
supposed to have the several parts of the grain present in differing proportions. 
