July 15, 1921 
Relation of Hardiness to Protein Content of Wheat 521 
The absence of correlation between hardness of the grain and per¬ 
centage of protein is remarkable, in view of the very general belief that 
the hard wheats are relatively high in protein, and the fact that this 
belief is supported by strong indirect evidence. 
SUMMARY 
(1) It has been assumed, and there is some evidence to show, that 
hardness of wheat is associated with and possibly dependent upon its 
protein content, the harder wheats being the higher in protein. 
(2) In the breeding of hard winter wheat, protein is one of the most 
important factors next to yield to be considered. This being the case, 
it is desirable to discover a method of selecting for protein that will dis¬ 
pense with the necessity of making chemical analyses of large numbers 
of varieties or strains under preliminary trial. 
(3) hardness and protein content are genetically related, then when 
large numbers are taken the coefficient of correlation should show such 
relationship. This being the case, the selection of strains for hardness 
should involve selection for protein also. 
(4) Three methods of testing wheat for hardness have been devised: 
That of the writer at the Kansas Agricultural Experiment Station (j); 
that of Harper and Peters, of the Kentucky Agricultural Experiment 
Station {2 ); and that of Shaw and Gaumnitz, of the California Agricul¬ 
tural Experiment Station ( 4 ). Calculating the correlation between the 
crushing or breaking point of the kernel in grams and the percentage of 
protein for the data from the three Stations, we have for the Kansas data 
a correlation coefficient of only 0.02 ±0.003, I° r the California data 
0.01 ±0.061, and for the Kentucky data 0.03 ±0.1545. These consist¬ 
ently negative results from the three sources seem inexplicable in the light 
of the generally accepted belief that the harder wheats are usually higher 
in protein. 
(5) The correlation coefficient between specific gravity and protein 
and between volume of the grain and protein were also determined. The 
correlation coefficient was found to be 0.05 ± in the first instance and 
zero in the second. 
(6) The writer is at a loss thus far to account for the conflict between 
the generally recognized fact, on the one hand, that the higher protein 
wheats are the harder wheats, and the entire absence of demonstrable 
correlation between protein content and hardness by means of the cor¬ 
relation coefficient, on the other. 
LITERATURE CITED 
(1) Baidey, C. H. 
1913. MINNESOTA WHEAT INVESTIGATIONS, SERIES I. MILUNG, BAKING, AND 
CHEMICAL, TESTS crop oE 1911. Minn. Agr. Exp. Sta. Bui. 131, 42 p., 
6 fig., map. 
(2) Harper, J. N., and Peters, A. M. 
1904. protein-content oe the wheat kerned. Ky. Agr. Exp. Sta. Bui. 113, 
12 p., 1 pi. 
