TESTIXG WHEAT FOB PROTEIN 6 
laboratories were reached : Research or experimental, mill, elevator, 
and commercial. 
Data obtained from this cooperative study are summarized in 
Table 1. A study of this table throws some interesting light upon 
the accuracy of making protein tests by different laboratories. 
Table 1. — Summary of crude protein results obtained at various laboratories 
on 12 samples of icheat 
Average... 
Maximum . 
Minimum- 
Range 
Sample number 
Per 
cent 
10.62 
11.00 
10.34 
0.66 
Per 
cent 
11.75 
12.14 
11.36 
0.78 
Per 
cent 
12.76 
13.16 
12.40 
0.76 
Per 
cent 
12.59 
13.00 
12.42 
0.58 
Per 
cent 
12.65 
13.11 
12.21 
0.90 
Per 
cent 
15.40 
16.04 
12.89 
13.15 
Per 
cent 
12. 26 
12.60 
11.43 
1.17 
Per 
cent 
12.78 
13.26 
12.04 
1.22 
Per 
cent 
14.30 
14.68 
13.46 
1.22 
Per 
cent 
10.40 
10.85 
10.00 
0.85 
Per 
cent 
10.55 
11.01 
10.26 
0.75 
Per 
cent 
11.07 
11.70 
10. 65 
1.05 
i This figure is believed to be an error in reporting. 
With one exception, the maximum difference found with any one 
sample was 1.22 per cent ; the minimum difference was 0.58 per cent. 
These figures are much larger than are usually associated with good 
laboratory technique in making protein tests, and are considerably 
larger than that allowed by either the Kansas City or the Omaha 
boards of trade. The} 7 allow 0.12 and 0.20 per cent, respectively. 
In the case of sample 6, one laboratory reported a test that varied 
3.15 per cent from the maximum test found by the other chemists. 
It is believed that there must have been a mistake in the calculations 
made by this laboratory, as the other anal} T ses made by this labora- 
tory- were within the ranges obtained by the other laboratories. 
This test will be considered as an error in reporting and will not be 
included in the discussion. 
There was a greater range in results with wheats of the hard red 
spring class than with wheats of the hard winter class. 
The size of the samples ground to make the protein test varied at 
the different laboratories from 6 to 60 grams. One collaborator did 
not grind his sample, but made the test on the whole grain. Roughly 
speaking. 65 per cent of the laboratories digested a 1-gram sample 
and 35 per cent digested a 2-gram sample. The quantity of acid 
used in the digestion mixture varied from 15 to 25 cubic centimeters 
with the laboratories using a 1-gram sample. Thirty per cent of 
those digesting a 1-gram sample used 25 cubic centimeters; 25 per 
cent, 15 to 18 cubic centimeters: and 38 per cent, 20 cubic centi- 
meters. Of the laboratories using a 2-gram sample for digestion, 40 
per cent used 25 cubic centimeters of acid ; 40 per cent, 30 cubic cen- 
timeters; and 18 per cent, 20 cubic centimeters. 
The catalytic agents used were numerous — sodium sulphate, po- 
tassium sulphate, metallic mercury, yellow oxide of mercury, red 
oxide of mercury, mercuric sulphate, metallic copper, and copper 
sulphate. Approximately 63 per cent of the laboratories used so- 
dium sulphate in quantities varying from 3 to 15 grams ; 35 per cent 
used potassium sulphate in the same quantities. Considering the 
mercuric catalysts, 18 per cent used metallic mercury; 10 per cent, 
