Per cec t. 
Invert sugar none 
Sugar trace 
Dextrin trace 
Starch i . i r 
Xylan, inverted by dilute acid 3 -76 
Xylan, soluble in alkali solution 0.15 
Lignones, rendered soluble by chlorin. ... 6.66 
Cellulose -5 • 59 
Moisture 7.21 
Ash Q.81 
Ether extract 1.15 
Proteids 15 • ib 
Soluble in alcohol 13 -^7 
Soluble in water (starch, etc., deducted) . 11.88 
Not determined 3*65 
100.00 
The substances dissolved by water are only partly pre- 
cipitated upon the addition of a large excess of alcohol. 
The amount precipitated was 8.2 per cent, of the sample. It 
did not prove to be dextrin. This caused me some trouble 
and 1 increased the amount of hydric chlorid used in invert- 
ing and increased the time of heating, at the temperature of 
boiling water, to one and a half and even to two hours. I 
did the same with the starch solution. 
A portion of the sample was extracted successively with 
95 per cent, alcohol and cold water — two grams of the sam- 
ple, 24 c. c. alcohol, and subsequently 40 c. c. cold water — 
the residue was washed with cold water and the remaining 
nitrogen determined. We found 1.554 percent., calculated 
on the air dried substance ; whereas the total nitrogen was 
equal to 2.426 per cent., a difference of .672 per cent., which 
is over 27 per cent, of the total nitrogen. This is rather more 
than we have found in any sample in the form of amid nit- 
rogen ; but inasmuch as some samples have furnished nearly 
as much as this, we are led to believe that it is the amids, 
principally if not wholly, which are soluble in water. There 
is in this no oxidation and fermentation such as take place 
in the weathering of hay, and the two processes are not 
equivalent, though they may be similar to some extent. 
Fhe crude fiber contained, in this instance, .33 per cent, 
nitrogen, or, roughly calculated on air dried hay, about .11 
per cent. This was neglected in the further calculations. 
Of the crude fiber itself 78.96 per cent, was cellulose, and 
21.04 per cent, was lignones. 
