— lO — 
results in showing the second cutting to be, in point of com- 
position, the preferable one. While these samples are the 
farthest removed from the general averages, they differ 
from these less than samples of the same cutting in any 
year may differ from one another. I am convinced that the 
variation in the different cuttings from year to year is de- 
pendent mostly upon the seasons, and is in no case very 
large, very much smaller in fact, than I had supposed. 
CHANGES IX COMPOSITION OE OLD HAV. 
That fresh hay, i. e., hay which has been in the mow or 
stack from one to nine months, is preferable to hay which 
has been there for a longer time, is generally conceded, 
whether there is any good ground for the general belief or 
not. The samples which we have used to study this ques- 
tion were prepared for analysis, put in glass bottles, and 
from one to two years allowed to elapse between the two 
determinations. These samples were stored in a dark cup- 
board in a dry room which was as close an imitation of the 
conditions prevailing in a mow as we could produce. The 
samples were air dried when put away, but they evidently 
changed in the amount of moisture present, as it was found 
necessary to redetermine the moisture in all of the samples. 
I expected that the crude protein would be most suscepti- 
ble to changes, and therefore determined upon making a 
series of nitrogen determinations from which to judge of 
the amount of deterioration. We found that we were 
wholly wrong in our idea that the nitrogen would be the 
most sensitive measure of any changes, at least for the con- 
ditions under which our samples were preserved. It may 
be noticed here that every sample had increased in the 
amount of moisture contained ; only one remained un- 
changed and this was the only sample which showed a 
diminution in the amount of nitrogen present. The 
changes, with this one exception, were all in the same di- 
rection— to an increase in the percentage of nitrogen. We 
examined seventeen samples and found but one exception 
to this rule. The bottles containing these samples were 
stoppered and sealed with paraffin, so it would seem very 
Improbable that the whole seventeen should fail to keep out 
the atmospheric moisture. We give the moisture determi- 
nations to show how marked this increase was and how 
general the rule, there being only one pronounced excep- 
tion to it. The first column contains the figures represent- 
ing the hay in 1894, the second In 1896. 
