gen to the soil and to render other plant food more avail- 
able. 
CRUDE FIBER AND NITROGEN-FREE EXTRAC'l'. 
Our subsequent analyses of these samples, of pea-vine 
hay, with the object of getting more definite information 
concerning the composition of the crude fiber and the nitro- 
gen-free extract, resulted as follows: 
In full bloom. 
Per cent. 
In full pod. 
Per cent. 
Invert sugar 
0.000 
0.000 
Cane sugar 
0.000 
3 . 050''^ 
Dextrin 
0.738 
0.705 
Starch 
0 . 000 
2.530 
Xylan, inverted by dilute acid. 
■ 8-157 
7-237 
Xylan, soluble in alkali soluthi 
0.816 
0.659 
Lignones, soluble by chlorin. . 
6.466 
10.296 
Cellulose 
18.646 
18.199 
Moisture ... 
5-^71 
6.028 
Ash 
11.273 
7-135 
Ether extract 
3 . 200 
I - S39 
Proteids 
20 . 200 
16.581 
Soluble in alcohol ( 
2S . 345 
25-8411 
Soluble in water \ 
Not determined 
1 . 258 
.000 
I oo . ooo 
100. 100 
I was quite surprised at the absence of starch in the 
sample in full bloom, but no more so than at the small 
amount of this substance in the alfalfa and the very small 
amount in the clover hay. The result for the alfalfa is en- 
tirely consonant with the results of the following experi- 
ment : A portion of the alfalfa was digested with alcohol 
and subsequently with cold water, to remove as much of 
the coloring matter and extractives as possible and then 
examined under the microscope, at last with the addition 
of a solution of iodin in potassic iodid, the reaction with 
the lignocelluloses was very strong, and if there was any 
reaction for starch it was entirely masked. I at no time 
succeeding in obtaining a satisfactory test for starch in this 
manner, though some of the tests might have been inter- 
preted as showing its presence, and I think this interpreta- 
tion is correct; but it was not clear enough at any time to 
indicate a large percentage of starch, even after boiling 
* A t?ecoii(J fi(-terniination jravp 2.^<4 pc*r cent. 
t Siifiar, dextrin, etc., deducted. 
