150 
which the pieces had taken up soda it was found that they 
had diminished in strength. On the other hand it was found 
that in such places of contact the silicate of soda of the sili- 
cated goods had suffered a partial decomposition, extending 
to the depth of four or five layers of the pieces. The silicate 
of soda in the middle of the pieces contained from 70 to 74 
per cent of silicic acid, combined with from 30 to 26 per cent 
of soda ; whilst the analysis of the silicate of soda contained 
in the contact layers showed that from one-third to two- 
thirds of its soda had been abstracted This loss of soda 
was accompanied by a change of strength of the cloth which 
appeared to bear some proportion to it ; the layers or folds 
of the cloth decreasing in strength as they were removed 
from contact with the starched goods until the silicate of soda 
attained the same composition as that found in the most 
rotten parts of the piece, this generally taking place about 
the fourth or fifth layer or fold of the piece, as before 
stated. 
The following table shows the changes in strength produced 
by this decomposition of the silicate of soda. 
Finished with Staech. 
Finished with Silicate of Soda. 
Middle, 
Contact Layer. 
I Contact Layer. 
2nd. 
3rd. 
4th.- 
5th. 
Middle. 
Strength ... 
100 
81 
89 
68 
62 
64 
48 
35 
The silicate of soda had evidently been decomposed with 
the formation of free alkali and an acid silicate which appears 
to have very little action on the cotton fibre. 
In some places the decomposition had gone further, and 
free silicic acid had separated out in the form of a white 
powder upon the surface of the cloth. The same decomposi- 
tion, accompanied by the same changes in the strength of the 
cloth, was observed upon all pieces which had been in con- 
tact with the paper used for wrapping the bales. In this 
