6o 
Colorado Experiment Station 
foods, wjiich we can recognize as specifically due to them, and are our 
analytical results accurate enough to justify us in assuming that the 
differences found are well enough established to be taken as the basis 
of inferences in regard to these effects? We should extend these ques¬ 
tions to include regularity in the direction of these differences, not 
only for the samples of the different varieties for one year, but in 
these same varieties for several years. 
I am very firmly convinced that minute differences as expressed 
in percentage composition correspond to very significant differences 
in the properties of the wheat,, provided that we can be certain that 
the percentages obtained are correct within limits, as narrow as or 
narrower than those with which the properties of the grain actually 
vary. I fear that it is so difficult to attain to this degree of certainty 
in regard to our analytical results, that these questions in regard to the 
specific effects of fertilizers and the manner in which they modify 
one another must remain in some doubt. 
The analyses given in the preceding tables are all based on the 
air-dried grain. The moisture in our samples is quite constant for the 
individual varieties grown on the same section of land. There is but 
one instance in nine groups of four samples each in which' we find a 
difference as great as or greater than i percent, so that the differences 
in the percntage, of potassium, for instance, due to this variation will 
fall in the third decimal place and wilH be less than five. The ash 
constituents in the wheat were determined by one analyst, and the 
phosphorus and potassium in the general analyses of wheat by another. 
The methods used in the case of the phosphorus were the same, but 
they used different methods in determining the potassium. The differ¬ 
ence in the percentages of the potassium due tO' this cause is from two 
to three one-hundredths of i percent. A comparison of the results 
obtained will show but one or two serious discrepancies. On the other 
hand, we find such agreement in the big features of the results that we 
can accept them as established on a pretty firm basis. 
There are reasons why we should prefer not too accept the results 
of the 1913 crop as indicative of the effects of the soil factors upon 
the composition of the wheat. One reason is that the land had been 
in other crops during the preceding years and we do not know with 
as much certainty as we should the conditions preceding, the growing 
of this crop. On the other hand, this crop grew and matured under 
favorable weather conditions, and probably gives us the most repre¬ 
sentative samples of our average wheat that we have obtained, but it 
is doubtful whether it gives us the best illustration of the influence of 
the soil factors upon the composition of the wheat. 
