— 19 — 
corresponding to 17.69 per cent. And in like manner for 25 
and 26, assuming 33.07 per cent., as the percentage of 
crude fiber in the stems, we find the xylan in 25 corre- 
sponding to 33.33 per cent, of the xylan in their original 
sample, and in 26 it amounts to 45.29 per cent. The method 
of preparation of the crude fiber was not the same in 
each case and these samples are not strictly comparable as 
far as the method affects the question. 
If we express these results in the more direct manner 
of stating the percentage of the total x^dan removed by the 
successive digestions with sulphuric acid and caustic soda, we 
find that in the case of the leaves 82.31 per cent, of the xylan 
was removed ; in the upper part of the stems, the small 
ends, we have 66.66 per cent, removed, and in the lower and 
more woody part we hnd that 54.71 per cent, has been re- 
moved. While it may not be rigorously correct that the 
amount of those substances which yield the furfurol are 
proportional to the furfurol obtained, it may be assumed to 
be true for the aggregate which we have in hay, and an ex- 
amination of the percentages of xylan shows that they, like 
the proteids, depend probably not so much upon differences 
of soil, as upon seasonal differences. Taking the three cut- 
tings of 1896, the second and third show an increase in the 
x^dan as the plant matures, but the first cutting shows the 
opposite. The samples collected in 1894 equally indefi- 
nite. W e are, however, justified by analyses 21 to 26, in 
concluding that there are several complexes present which 
yield furfurol, and they offer different degrees of resistance 
to the alternate action of acids and alkalies. The com- 
plexes yielding more readih' to the action of these agents 
predominate in the leaves, forming nearly nine-tenths of 
the whole amount, whereas in the stems they form only 
about one-half. 
Analyses to be given subsequently indicate that there 
is still another distinction, for some of them are removed by 
extraction with alcohol and water, while others are not. 
The conditions, under which the work recorded in these 
paragraphs was made, were as uniform as possible, for it is 
evident that this is necessary’ in order that our results shall 
have the same significance. 
COMPARISON OF LK(iUMIXOUS IIAVS I!V NKW M l-'/l'lIO 1 )S . 
It has long been felt that our methods of fodder analy- 
sis leave much to be desired, and, while they are very help- 
ful in forming a judgment of the value of a hay, they have 
not been conclusive. I do not claim that the present effort 
