238 Proceedings of Royal Society of Edinburgh. [sess. 
that silicate of magnesia, in an amorphous condition, is distinctly 
soluble in pure water, and also to a greater extent in sea-water. 
To determine this we added pure silicate of soda to sea-water in 
the following proportions : — 
Silicate of soda representing — 
part silicic acid added to 1,000 parts sea-water produced a large precipitate. 
„ gave a distinct precipitate 
10,000 
20,000 
30.000 
40.000 
50.000 
100,000 
precipitate after 24 hours. 
36 ,, 
48 „ 
144 ,, 
144 . 
These precipitates consisted principally of silicate of magnesia, 
silicic acid, and traces of silicate of lime ; so that we are justified in 
concluding that if alkaline silicates in proportion equal to 1 part of 
silicic acid in 100,000 parts of water were present, the whole would 
be removed in combination with magnesia and thrown out of solution. 
For the sake of argument, let us suppose that silicic acid did occur 
to this extent in sea- water, thus representing 0 ’01 grm. per litre, even 
this amount, taking it, we shall say, as a saturated solution of the most 
insoluble silicate known, is very much less than the amount of silicic 
acid found by the analysts whose results represent the maximum 
amount of that body found in sea-water (see Table I., p. 235). 
These experiments were repeated with pure water, to wdiich chloride 
of magnesium was added together with pure silicate of soda. The 
results showed that amorphous or freshly-precipitated silicate of 
magnesia was less soluble in fresh than in sea water, thus leading 
us to assume that it could not be carried to the sea as sush by rivers, 
as also that river water, rich in other magnesium salts, can hardly 
be supposed to carry soluble silicates to the sea. 
A similar series of experiments were conducted with lime salts, 
but the amorphous silicate of lime so formed was found, in com- 
parison with the silicate of magnesium, to be so soluble that it did 
not cause any precipitate above 1 part in 50,000 of water in 
twenty-four hours. 
To vary^ these experiments, a weak solution of silicate of soda 
w r as exactly neutralised by means of hydrochloric acid ; an amount 
representing 1 grm. of soluble or colloid silicic acid was added to 
