samples from other states, a great uniformity in composi- 
tion. The crude protein in the first cutting hay had a range 
of less than two per cent. — from 14 to 15.9 per cent. These 
samples represented hay made from alfalfa grown with and 
also without irrigation ; also such as had been grown upon 
high land as well as that which had been grown upon low 
land. The crude fiber in these samples had a very much 
greater range in percentage — from 32 to 40 percent. — than 
the crude proteids. This may have been partly due to a 
less degree of sharpness in the method of determination, 
but we judge it to be due to an actual difference in the sam- 
ples. These are narrow limits when compared with those 
shown by analyses representing different states, in which 
we have, for the crude protein, a range of 14.6 per cent.— 
from 1 1. 1 to 25.7 per cent.; and for the crude fiber, a range 
of 24.5 per cent., or from 15.4 to 39.97 per cent. 
In the second cutting again we find a narrow range, but 
wider than in the first cutting. The extreme range for the 
crude protein is 3.6 per cent., and, excluding an exceptional 
sample with 18.47 cent., this range becomes 3.5 per cent. 
The crude fiber for this cutting has a range of about 12 per 
cent. — from 26.16 to 38.08 percent. 
The range for the proteids in the third cutting is 3.5 per 
cent., and for the crude fiber, 10 per cent, riiese samples 
were all of hay as put into the mow or stack. 
A study of the analyses published up to 1896, fails to 
furnish any general composition for this fodder, and we 
cannot discern any patent and adequate reason for this. In 
the New Jersey Experiment Station Report, for 1886, the 
lowest amount given for crude protein was 16 per cent., and 
the highest for crude fiber was 35 per cent.; as given in the 
Report for 1888, eight samples are recorded ; the lowest of 
these in proteids has 13.24 per cent., and the highest in 
crude fiber has only 24.34 per cent. The same observations 
are true of the Texas samples, except that we find a greater 
difference between the highest and lowest in the case of 
both of these constituents ; for the proteids, from 13.31 'to 
25*75 per cent., and for the crude fiber, from 16.64 to 34.23 
per cent. If we take the published analyses of Colorado 
samples we shall find a like variation. 
lAen the most unfriendly critic of the methods or the 
operators using them, cannot possibly explain these differ- 
ences by the weaknesses of either or both of these. In 
order to obtain light upon this point, we have studied the 
three cuttings of 1894, the first cutting of 1893, the three 
cuttings of 1896. The methods employed in the anal3'ses 
