A Study ot~Cotorado Wheat 
99 
kernels, or 44 percent in favor of the 1913 crop. The difference in the 
average percentage of starch, taking all the samples except those grown 
with nitrates, is 0.84 percent. One hundred grams of the shrunken 
wheat contained practically the same amount of starch as a like quantity 
of the good wheat, and it contained less protein by 3.58 grams. This 
would seem to indicate that deposition of nitrogen and starch takes 
place at the same time—or if the deposition of one was hindered more 
than the other, the deposition of the nitrogen was hindered more than 
that of the starch. 
VARIATION IN AMOUNT OF CRUDE FIBER 
No feature in the composition of the crops produced during these 
three seasons is more marked in its variation than the percentage of 
crude fiber. In 1913 it was, perhaps, a little below the average for 
wheat in general—it may be taken roughly at 2.60 percent for 1913 — 
in 1914 it was higher, about 3.00 percent, and in 1915 it was very high, 
about 3.7 percent. It will be observed that this is roughly inversely 
proportional to the absolute weight of the wheat. In other words, the 
frame-work of the cells had been completed and probably but little or 
no cellulose would under any conditions have been added to the kernels 
subsequent to the period at which the rust attacked the plants. The 
rust stopped the transfer of the filling material from the plainjt to the 
. kernels, and left the ratio of cellulose to the other constituents abnor¬ 
mally high and the kernels shrunken. 
The ash, as determined, is probably about o.i percent too low, 
owing to the losses in making the determination, but this loss is com¬ 
mon to all the determinations and in the same direction, so that the 
averages are not very much disturbed. The individual determinations 
are comparatively close together, ranging in the case of Section 1800 
from 1.700 to 1.995 for the year 1913, from 1-692 to 2.017 I9i4> 
and from 1.777 to 2.022 in 1915. These data include* the three varie¬ 
ties and the different fertilizations. The averages for the three years, 
1913, 1914 and 1915 are, for the grain grown on Section 1800, 1.925, 
1.817 and 1.778 percent of ash respectively. The statement of the ash 
constituents has been given in percentages of the air-dried grain in¬ 
stead of percentages of the ash, which reduces the numbers given in 
the statement- Disregarding the chlorin and sulphur in the present 
statement, the greatest variations will be found in the phosphorus and 
potassium. These two are perhaps the most important ash constituents, 
at least they are in our problem, so far as now developed. The average 
amount of phosphorus in the grain (all three varieties taken) grown 
on Section 1800 in 1913, 1914, and 1915, was 0.432, 0.388 and 0.382 
percent respectively, and the average amount of potassium was 0.441, 
0.448 and 0.426 percent. While these averages conceal the differences 
