figures of the bottom layer analysis, adding these to the
second and third layer analysis, and dividing by 4, we obtain a fair
representation of the average composition of the silage taken
throughout the silo, for by so doing we obtain the average of the
analyses of each 6-inch layer of silage. The results of the analyses
are as follows, calculated on the dry matter. The moisture was
practically the same, being 70.48 per cent, in the grass and 72.97 in
the silage.

_Composition of Grass and Silage (dried at 100 deg.C.)._

Grass. Ensilage.
Fat (ether extract) 2.80 5.38
Soluble albuminous compounds 3.06 5.98
Insoluble albuminous compounds 6.94 3.77
Mucilage, sugar, and extractives, etc. 11.65 4.98
Digestible fiber 36.24 33.37
Indigestible woody fiber 32.33 31.79
------- -------
93.02 85.27
Soluble mineral matters 5.24 12.62
Insoluble mineral matters 1.74 2.11
------- -------
100.00 100.00

The striking difference in the mineral matter of the grass and silage
I will merely draw attention to; it is not due to the salt added to
the silage. I may say, however, that other analysts and I myself have
found similar striking differences. For instance, Prof. Kinch[2]
found in grass 8.50 per cent. mineral matter, in silage 10.10 per
cent., which, as be points out, is equivalent, to a "loss of about 18
per cent. of combustible constituents"--a loss which we have no proof
of having taken place. In Mr. Smetham's sample the loss would have to
be 50 per cent., which did not occur, and in fact is not possible.
What is the explanation?

[Footnote 2: _Journ. Chem. Society_, March, 1884, p. 124.]

I am, however, considering now the organic constituents. Calculating
the percentages of these in the grass and silage, we obtain the
following figures:

_Percentage Composition of Organic Compounds._

Grass. Ensilage.
Fat (ether extract) 3.01 6.31
Soluble albuminous compounds 8.29}