240 Proceedings of Boycd Society of Edinburgh. 
400,000 parts of water — figures curiously corresponding to those 
representing the minimum determinations of soluble silica in sea- 
water by us (see Table I.).* 
As was stated above, there are no means at our disposal to ex- 
plain the elaboration of silica from salts by organisms, as in 
the case of the secretion of carbonate of lime from other calcium 
salts, f Silicic acid, if present at all in the ocean in a soluble form, 
can only occur in that one condition. Where then are we to look 
for the sources from which Diatoms, Radiolarians, Sponges, &c., 
obtain the silica necessary for their siliceous skeletons ? 
Early in the course of these investigations we were led to suspect 
that the pelagic siliceous organisms might, in part at least, obtain the 
silica for their frustules and skeletons from the clayey matter sus- 
pended mechanically in sea-water. It has been long known that the 
principal part of the fine clayey matter suspended in river water is 
* Kyle, in his analyses of the water supplied to the city of Buenos Ayres, 
states that “the river Plate is in reality the estuary of the rivers Parana and 
Uruguay. It is characterised by its muddy appearance, and always contains 
in suspension a considerable amount of coloured clay.” In the analyses which 
follow of the waters of the rivers Plate and Parana, this clay is represented as 
alumina and silica. From this it is evident that the waters were not filtered 
before analysis. On the other hand, looking at the analyses of the water of 
the river Uruguay, the analyst characterises it “as a very remarkable one, and 
probably one of the purest river waters in the world, containing rather less 
than four parts of solid matter per 100,000. Alumina is entirely absent, the 
noteworthy fact being that about 46 per cent, of the total solid matter consists 
of soluble silica not suspended as in the other two rivers. A small proportion 
exists probably as alkaline silicates, but the greater part is undoubtedly present 
as hydrated silicic acid.” In these circumstances maybe found an explanation 
of the petrifying properties attributed to the water of the Uruguay ( Chemical 
News , vol. xxxviii. p. 28). 
The abnormal quantity of free silicic acid present in these waters may be 
accounted for, either by the decomposition of felspar rocks by carbonic acid, or 
by the action of azo-humic acids referred to by Julien (“On the Geological 
Action of Humus Acids,” Proc. American Ass., vol. xxviii. p. 325) on silica 
itself. But, granted all that has been advanced as to the carriage of silicic 
acid as such in a soluble condition to the sea by rivers, as we have shown, when 
such water mixes with the sea there can be no possible accumulation of soluble 
silica over that of one part in from 50,000 to 100,000 parts of water. 
Julien, we think wrongly, urges that the humus acids may have the same 
action in sea- water that they have upon silica on land in the presence of fresh 
water ; that a vast proportion of humus acids reaches the sea is undoubtedly the 
case, but immediately on mixing with salt-w T ater the humus acid is thrown 
down either in combination with lime, magnesia, or alumina. 
t See Murray and Irvine, “Coral Beefs and other Carbonate of Lime Forma- 
tions in Modern Seas,” Proc. Poy. Soc. Eclin., vol. xvii. pp. 79-109, 1890. 
