1890-91.] Dr Murray and Mr Irvine on Silica in Seas. 
237 
Even when any loss of silicic acid which may occur from this 
source of error is taken into account, the amount of soluble silica 
in sea-water is too small to be quite certain that it is sufficient to 
supply all that is required by silica-secreting organisms. 
Moreover, if we accept the fact that the sea receives the greater 
part of its silicic acid from the waters carried down to it by rivers 
in the form of soluble silicates, we should expect that such silicates 
would be decomposed by the salts of magnesium so abundantly pre- 
sent in sea- water, with the formation of silicate of magnesia. This 
substance we know, in the condition of steatite or talc, is so in- 
soluble, that no appreciable amount is removed by solution in water, 
even in presence of free carbonic acid.* We have found, however, 
the amount so added being 0 - 0342 grm., adding also 0'096 grm. of fluoride of 
sodium. The whole was then evaporated to dryness with hydrochloric acid, 
and the' residue dried and drenched with hydrochloric acid, and again dried. 
The silicic acid determined in the insoluble residue showed that 80 ‘4 per cent, 
of the whole silicic acid added had been thus recovered, only 19 ‘6 per cent, 
being lost by the action of the fluoride. A similar experiment (B) was conducted 
with the same amount of silicic acid as in Experiment (A), but 0 - 960 grm., or 
ten times the amount of fluoride of sodium, was added to the same amount of 
artificial sea-water. This quantity gave a precipitate of fluoride of calcium 
with the lime salts of the sea-water, so that such an amount of fluorine could 
not be present in natural sea-water. The amount of silicic acid recovered in 
this instance was 67 '5 per cent., the loss of silicic acid being 32 - 5 per cent. 
A third experiment (C) was made, in which a large excess of fluoride of sodium 
was again added, and the evaporation and drenching with hydrochloric acid 
repeated six or seven times in succession. It was then found that only 4 per 
cent, of the silicic acid added remained in the insoluble residue. These results 
would seem to show that even if sea-water contained as much silicic acid as 
one part in 50,000, and also as much fluoride as it could hold in solution, at 
least two-thirds of the silicic acid present in sea-water would be found by the 
methods ordinarily in use for silicic acid determination. Nothing like that 
quantity (1 in 50,000 parts) was ever found in our experiments. The third 
experiment (C) shows that the silicic acid may be carried away in a volatile 
condition, combined with fluorine, after repeated evaporations with hydro- 
chloric acid and ignitings, so that the amount of silicic acid we have found in 
carefully-filtered sea-water must be correct within the limits of at least 20 per 
cent, (see Experiment A), thus making it according to these determinations 
less than one part in 250,000. 
* Bischoff, Chemical and Physical Geology , vol. i. p. 3. Mr Alexander 
Johnstone, F.G.S., has proved experimentally that pure water, even 
when saturated with carbonic acid, has no solvent action on pure talc 
or steatite, but that sea-water has a slight but distinct effect in this direc- 
tion equal to 1 part in 200,000. His results are also interesting as show- 
ing that silicate of magnesium once formed cannot be conveyed to any extent 
in a soluble condition by river water to the sea ( Proc . Roy. Soc. Edin . , vol. 
xvi. pp. 172-175, 1889). ' 
