TESTING WHEAT FOE PROTEIN 31 
in protein content of as much as 0.3 per cent will result if smaller 
samples are used. 
From an expense standpoint, provided heat conditions are op- 
timum, the Gunning method can perhaps be used more economically 
than either of the others, other conditions being equal. 
Yellow and red oxides of mercury are equivalent in all quan- 
tities as catalytic agents; 0.5 gram gives satisfactory results. 
Mercuric sulphate gives slightly lower results than the other two 
salts studied. 
Sodium sulphate and potassium sulphate are equal as catalytic 
agents. Ten grams of either give best results when 20 cubic 
centimeters of acid are used. Sodium sulphate, however, has the 
disadvantage of forming a salt cake with the usual quantities of 
sulphuric acid used for digestion. This salt cake can be prevented 
by using a mixture of 60 per cent K 2 S0 4 and 40 per cent Na 2 S0 4 . 
Copper sulphate as a catalytic agent is of little value at low 
intensities of heat. 
Wheat should be ground whenever possible. Protein results from 
samples of whole kernels are irregular when compared to data ob- 
tained from carefully ground samples. 
There is no difference between electricity and gas as sources of 
heat, electricity having the advantage in most instances of giving 
a greater intensity of heat. 
Twenty cubic centimeters of concentrated sulphuric acid are neces- 
sary to digest and retain all the ammonia liberated from a 1-gram 
sample of wheat at the intensity of heat usually employed by most 
laboratories. Twenty-five cubic centimeters are necessary when a 
2-gram sample is used. 
A 2-gram sample gives slightly higher results than a 1-gram sam- 
ple. From an error standpoint, the advantage of using a 2-gram 
sample is not great enough to warrant its use. 
Considering the extra time needed to digest a 2-gram sample, 
the 2-gram sample is undesirable. 
The Kjeldahl method is the only method with which wheat sam- 
ples can be completely oxidized in less than an hour at all intensi- 
ties of heat. The Gunning method and the Kansas City Protein 
Referee Board methods are not satisfactory at low heats. 
No difference in results can be attributed to methods if the heat 
is sufficiently intense. Wheat can be completely oxidized in 45 
minutes at high heat arid in 60 minutes at medium heat, when a 
1-gram sample is used. It takes 30 per cent more time to digest a 
2-gram sample. 
If reductions in the quantities of catalytic agents are made from 
the figures given in the formulas studied, they will have to be com- 
pensated for by more intense or longer heat. 
In the distillation process careful attention should be given to the 
use of traps and to the making of blank determinations. 
At least 150 cubic centimeters of distillate should be collected 
in acid which will hold the equivalent of 35 to 70 milligrams of 
nitrogen, depending upon whether a 1 or 2 gram sample is used. 
There is no choice between sulphuric and hydrochloric as re- 
ceiving acids. " 
Boric acid will lose ammonia at 40° C, the loss increasing with the 
rise in temperature. 
