Vol. 51.] THE CHEMICAL COMPOSITION OF OCEANIC DEPOSITS. 321 



A. B. C. D. 



Loss on ignition 667 8"55 472 710 



Quartz '01 41 -73 76 



Colloid silica 4-61 498 576 4-85 



Combined silica 49"48 4765 5450 5170 



Peroxide of iron 10-55 4-86 695 2-84 



Alumina 1571 25'20 2103 25-48 



Peroxide of manganese... 156 3 - 05 124 2 - 71 



Lime .. 101 trace — — 



Magnesia 1-38 257 2"62 268 



Potash 1-88 -18 -50 -10 



Soda 322 255 1-95 178 



Carbonate of lime "92 



Sodium chloride L84 



Magnesium chloride ... "96 



Sulphate of lime "80 



100-60 10000 10000 10000 



It will be seen that the chief differences are that the recent clay 

 contains sea-salts, a little calcium carbonate, less alumina and more 

 peroxide of iron than the Tertiary deep-sea clays. But if such a 

 recent clay were exposed to the action of percolating water, the sea- 

 salts, the calcium carbonate, and some of the iron would be dissolved 

 out, with the result that the proportions of silica and alumina 

 would be increased. If all the sea-salts, the carbonate of lime, and 

 half of the iron were dissolved from the clay A, the remainder 

 would contain 17*6 of alumina and 5o*44 of combined silica. 



There is still a difference between these amounts and those of the 

 Barbados clays, for none of the latter contained quite so much silica 

 or quite so little alumina. Still we found that the proportions of 

 silica (combined) varied from 43 to 54*5, and those of alumina from 

 21 to 31 - 5, the amount of alumina thus being always large. 



We believe that all these deep-sea clays are mainly formed by the 

 suboceanic decomposition of volcanic glasses and pumices, and that 

 the variations in the amounts of silica, alumina, and alkalies which 

 they contain depend on the relative proportions of the basic and acid 

 varieties of these materials. From the large proportion of alumina 

 in the Barbadian clays it would appear that they have been derived 

 from the decomposition of very basic materials, while the high per- 

 centage of silica in many recent deep-sea clays is evidently due to 

 the presence of a highly acidic pumice. 



In this connexion it is noticeable that the amount of potash is 

 very small in the Barbadian clays, while that of soda is compara- 

 tively large. In this and other respects the similarity between the 

 recent clay and the Tertiary clays is very striking, the minute quan- 

 tity of quartz and the constant 4 or 5 per cent, of colloid (organic) 

 silica being essential characteristics of deep-sea clay, whether ancient 

 or modern. 



As regards the composition of the Barbadian clays, we find that 

 if the pumiceous dust which actually fell on Barbados in 1812 from 

 the eruption of St. Vincent 1 were mixed with a clay such as we 



1 See analysis in our paper on the Oceanic Deposits of Barbados, Quart. 

 Journ. Geol. Soc. vol. xlviii. (1892) p. 192. 



2a2 



