84 
Colorado Experiment Station 
SHRUNKEN KERNELS NOT NECESSARILY HIGH IN PROTEIN 
In the preceding tables we have the samples grown during two 
y’ears of experimentation. Many of these samples, practically all those 
of 1915, the Defiance samples of 1914, and the samples of both the 
Fife and Kubanka grown with the application of nitrate are shrunken. 
These latter samples are not shrunken tO’ anything like the extent of 
some of the others. I think that it is generally accepted as a fact that 
shrunken grains are higher than normal in protein. I acknowledge 
that I thought that this was an established fact. But there is no differ¬ 
ence between the badly shrunken samples and the plumper ones that 
cannot justly be explained in some other manner. The samples grown 
with the application of nitrates in 1913, carried^ in round numbers, 2.0 
percent more protein than samples from the other plots. We find on an 
average a little greater difference in 1915. Our samples of 1913 are all 
fine, plump wheats, as the weights per 1,000 kernels show. In 1915 
thev were all more or less shrunken. The Defiance was so badlv 
shrunken that we could sell it only for chicken feed. It would seem 
unreasonable to consider the excess of protein in the nitrate samples of 
1915 due to their shrunken condition, when essentially the same differ¬ 
ence existed in the 1913 crop, which was not shrunken. The other 
shrunken samples of 1915 are all low in protein but they are neither 
higher nor lower than the less shrunken samples of the same variety 
grown that season. In fact, the very badly shrunken samples of 1915, 
differ but little in their protein content from the plump, full-weight 
samples of 1914. The difference is in favor of the plump samples of 
1914. The fact is that the shrunken condition of the samples seems to 
have nothing to do with the composition, but simply indicates that the 
process of filling was cut short, and that which would have been a 
large kernel, remained an incompleted structure. The character of the 
material, and the amount of it, with which the kernel was filled, was in 
these cases determined by the rust. The similarity in the composition 
of the crops of 1914 and 1915 is remarkable. Rust was very abundant 
in both seasons. The plants were a little more fully developed in 1914 
than in 1915 at the time that the rust'attacked the plants. The crop 
was materially better iiT 1914 than in 1915, but the composition of the 
wheat was, as stated, very similar. It is not to be understood that we 
had no fair crops in 1915, for this is not the case. We had as high as 
40.8 bushels per acre, weighing 60 pounds to the bushel, and it was 
only the plots treated with nitrates that yielded crops of either Red 
Fife or Kubanka that weighed (less than 60 pounds to the bushel, some 
of the Kubanka weighed 63 pounds, but the composition of the wheat 
was essentially the same as that of the badly shrunken Defiance, weigh¬ 
ing from 49 to 53 pounds per bushel. 
