TESTING WHEAT FOR PROTEIN 
17 
TAKING OUT THE SUBSAMPLE 
Before a subsample is taken, the original bulk should be thor- 
oughly mixed because of the influence which sizes and types of 
wheat kernel have on the protein test results. It is common knowl- 
edge that the dark, hard, and vitreous kernels of wheat in any given 
sample have more protein than the soft and chalky kernels. This 
has been brought out by Frank lx and is supported by the data in 
Table 11. Differences in protein content of dark, hard, and vitreous 
kernels as compared to chalky kernels will average about 1.5 per cent 
and cases are on record in which the differences were as great as 4 
per cent. If the quantities of dark, hard, and vitreous and chalky 
kernels vary appreciably, differences in results are bound to occur 
unless the samples are thoroughly mixed and reduced by use of a 
sampling device. Therefore one can not be too careful in getting 
a representative sample of the grain to be analyzed before making 
the tests. 
Table 11. — Protein content of yellow and chalky and dark, hard, and vitreous 
portions of hard red winter and hard red spring ivheats 
Number of samples 
Average protein content per cent.. 
Maximum protein content do 
Minimum protein content do 
Range in protein content do 
Hard red winter 
wheats 
Dark, hard 
and 
vitreous 
156 
12.34 
16.24 
8.82 
7.42 
Chalky 
156 
11.11 
15.21 
7.40 
7.81 
Hard red spring 
wheats 
Dark, hard, 
and 
vitreous 
70 
13.15 
16.24 
11.58 
4.66 
Chalky 
70 
11.42 
14.31 
9.74 
4.57 
CORRECT SIZE OF SAMPLE TO GRIND 
From the collaborative data it was found that the quantity of 
wheat ground varied from 6 to 150 grams. To determine how great 
an effect the grinding of small and large quantities would have on 
the protein test results, 1,000-gram portions of each of the 12 sam- 
ples submitted for collaborative tests were mixed thoroughly and 
passed through the Boerner sampler and 8 to 15, 30, 60, and 125 
gram portions were drawn off, ground, and protein tests made 
thereon. 
The data given in part A of Table 12 record the protein percen- 
tages without taking into consideration the moisture content of the 
sample. As much as 3 per cent of moisture may be lost in the grind- 
ing process. Differences of as much as 0.4 per cent of protein were 
obtained from the results of grinding varying quantities of wheat. 
The greatest differences were obtained with samples of 8 and 15 
11 Frank, W. L. protein content vs. texture vs. kernel weight in southwestern 
market-run winter wheat. Modern Miller, July 14, 1923, p. 23. 
