TESTING WHEAT FOR PROTEIN 
19 
mill. A second sample of low protein wheat, in 5, 10, 15, 20, and 25 
gram portions, was then passed through and the protein test was 
made on each grinding. After the 25-gram portion was run through, 
the mill was cleaned and a second 25-gram portion of the sample was 
cleaned with 5, 10, 15, 20, and 25 gram portions of the high protein 
wheat. The experiment was repeated with second high and low pro- 
tein wheat. 
It is evident that from 15 to 20 grams of wheat are sufficient to 
clear the mill for any new samples. A small quantity of high or low 
protein wheat left in the burrs will cause erroneous results. 
Table 13. — Grams of tclieat necessary to clean small-burr mill in order to 
obtain correct results 
Treatment 
Protein 
Treatment 
Protein 
SERIES I 
Sample 12003.. 
Per cent 
15.45 
12.25 
10.94 
10.72 
10.72 
10.72 
10.72 
13.39 
15.28 
15.45 
15.50 
15.45 
series m 
Sample 12007 
Per cent 
10.26 
Cleaned with 5 grams 11998 ... 
C lea ned with 5 grams 12006 _ 
13.22 
Cleaned with 10 grams 11998. 
Cleaned with 10 grams 12006 
14.36 
Cleaned with 15 grams 11998 . 
Cleaned with 1 5 grams 12006 _ 
14.45 
Cleaned with 20 grams 11998. 
Cleaned with 20 grams 12006 
14.48 
Cleaned with 25 grams 11998 
Cleaned with 25 grams 12006 
SERIES IV 
Sample 12006. 
14.50 
series n 
Sample 11998. 
14. 50 
Cleaned with 5 grams 12003 
Cleaned with 5 grams 12007 
11.80 
Cleaned with 10 grams 12003 
Cleaned with 10 grams 12007. 
10.32 
Cleaned with 15 grams 12003 
Cleaned with 15 grams 12007 
10.20 
Cleaned with 20 grams 12003 
Cleaned with 20 grams 12007 
10. 25 
Cleaned with 25 grams 12003 
CM pan p.d with 25 grams 12007 
10.29 
INFLUENCE OF FINENESS OF DIVISION OF SAMPLE ON PROTEIN-TEST RESULTS 
The fact that little difficulty is experienced in getting concordant 
results when making protein determination on flour is evidence that 
fineness of division in grinding may have a bearing on the accuracy 
of protein tests in wheat. 
To study this point, 100-gram portions of the same 12 samples 
of wheat before described were analyzed in five different states of 
division, ranging from the whole seed to that condition where 100 
per cent of the ground sample passed through a No. 36 grits gauze. 
The fineness of division series in detail were: Whole kernel, kernels 
ground so that 15, 30, 60 and 100 per cent, respectively, passed 
through a No. 36 grits gauze. Before analysis, the samples were 
carefully mixed on paper to insure uniformity. 
Data on this subject are given in Table 14. Part A records the 
analyses in which no consideration was given to the moisture con- 
tent, and part B the data on a 13.5 per cent moisture basis. 
It is evident from these tests that the fineness to which a sample 
is ground influences the accuracy of protein tests. This is doubly 
true when moisture is not considered. The maximum difference 
under such conditions, as shown in Table 14, part A, was found to 
be 0.45 per cent. The results with whole kernels are irregular. 
Sometimes they are high and at other times low, as compared with 
the results from carefully ground samples, 
