324 COMPOSITION AND NUTRITIVE VALUE OF COTTON-CAKE. 
On comparing the composition of these cakes with the 
average composition of decorticated cake, it will be seen : 
1. That the proportion of oil in the common cake is very 
much smaller than in the decorticated cake. The amount of 
oil is in reality inconsiderable, since most other oil-cakes, 
such as linseed, rape, poppy, and others, contain a good deal 
more fattv matter. 
2. In decorticated cotton-cake there is a much higher per¬ 
centage of albuminous compounds. 
3. On the other hand, the proportion of indigestible fibre 
in ordinary cotton-cake is far more considerable than in 
decorticated cake. 
Common cotton-cake is sold at a lower price than the de¬ 
corticated cake, hut, considering the inferior character of the 
former, its lower price offers little inducement to intending 
purchasers. A ton of the best decorticated cake, I am in¬ 
clined to think, is worth quite as much as two tons of the 
inferior cake made of the whole seed. I consider the latter 
dear at the price at which it is sold. 
The first sample of the three contained much more woody 
fibre than the two others. 1 may observe, however, that the 
proportion of woody fibre mentioned in the analyses of 
these cakes does not indicate the amount of seed-shells. 
Wh at is here called indigestible woody fibre was obtained by 
treating the cake with cold and boiling; water, and subse¬ 
quently with dilute potash solution and dilute sulphuric acid. 
By these means the albuminous compounds, pectinous 
matters, and other constituents of the seed-shells are ren¬ 
dered soluble, and the insoluble residue is conceived to be 
indigestible cellular fibre. 
I have endeavoured, approximately, to determine the pro¬ 
portion of shells in the first cake. As the shells are reduced 
to powder only with great difficulty, whereas the kernel 
powders readily enough, the proportion of the former may 
be ascertained with tolerable accuracy by carefully grinding 
the cake in a stone mortar, and sifting off the hard shells. 
Proceeding in this way, I found in No. 1, .58*42 per cent, of 
shells. This is a great deal more than the average amount 
of seed-shells in cotton-cake made of the whole seed. It 
appears to me, therefore, not unlikely that this cake con¬ 
tained an additional quantity of shells beyond that which 
cotton-seed in its raw state naturally furnishes. In the 
manufacture of decorticated cake large quantities of shells 
must be obtained, and it is not unlikely that these shells, 
