1890-91.] Dr Murray and Mr Irvine on Silica in Seas. 
239 
1 litre of sea-water, the one-half of which was kept at a temperature 
of 40° and the other at 80° Fahr. Here apparently more silicic acid 
was retained in solution by the sea-water than in the former experi- 
ment with the alkaline silicate, and on determining the amount of 
silicic acid in the filtered sea- water after seventy-two hours, it was 
found that sea-water could hold up for that period 0*1346 grm. per 
litre, the precipitate consisting principally of silicic acid and silicate 
of magnesia. We think, however, that as a rule silicic acid must 
reach the ocean in the form of silicates.* 
At this point it may he well to refer to the character of river 
water in its relation to the amount of silicic acid carried to the sea. 
From the analyses of forty-eight river waters given by Bischoff,f we 
find the average quantity of silicic acid present to amount to about 1 
part in 100,000. But here the same source of error seems to have 
crept in, as is the case with the silicic acid determinations in sea- 
water, a certain number of analysts showing a maximum amount 
equal to 4*8 parts in 100,000, whilst others show minimum results 
equivalent to 0*01 part in 100,000. That this source of error is due 
to the presence of insoluble clay, is pretty well proved by the results 
of certain of the analyses. For instance, in those of the waters of 
the Maas, the amount of silica at Hocht is 2 parts in 100,000 * at 
Pierrebleue, 1*04; and at Arensdonck, 0*28; clearly showing that 
the water = had deposited insoluble siliceous matter or clay in its 
course to the sea. However this may be, it is certainly curious to 
note, taking the minimum results of these forty-eight analyses as 
representing the true amount of soluble silicic acid in river water, 
that we find it equivalent to 1 part of that body in 250,000 to 
* A solution of silicate of soda was neutralised with carbonic acid, and an 
amount of this solution equal to 1 grm. of soluble or colloid silicic acid added 
to 1 litre of sea- water. A comparatively small precipitate resulted, and was 
found to consist of silicic acid and silicate of magnesia, with traces of lime. 
The clear liquid filtrate from this precipitate remained for a very long period 
perfectly clear, and only deposited a slight additional precipitate after stand- 
ing more than fourteen days. The silicic acid was determined in the filtrate, 
and it was found that the amount of the precipitate corresponded with the 
carbonate of magnesia or lime which occurred in the water. Thus if soluble 
silicic acid in any circumstances be added to sea-water, we should expect only 
that portion thrown out that would thus combine with the alkaline constituents 
of the sea-water, the amount of alkaline constituents being always enormously 
in excess over that in which silicic acid could exist either in surface or bottom 
water. t Loc. cit,, vol. i. pp. 76, 77. 
