224 CLARKE— SOME GEOCHEMICAL STATISTICS. [April 20. 



Igneous E. Sedimentaries. 



NaaO 3.49 1,21 



K:;0 3.07 3.03 



TiO. yj ■ .62 



PsOo 27 .14 



MnO 10 .05 



100.00 100.00 



This comparison shows an increase in siHca, due to concentration 

 Alumina is lowered, iron is nearly constant although the proportions 

 of the two oxides are reversed, and potassium is constant also. Soda 

 and magnesia show losses due to leaching. Lime is increased, 

 which suggests that the 5 per cent, of the sedimentaries assigned to 

 limestone is probably too high. Mead's estimate of 9 per cent, is 

 certainly excessive. The largest uncertainty is due to the facts that 

 the composite shale used for analysis did not include certain mag- 

 nesian rocks, and the limestones did not embrace the dolomites. The 

 deficiency of magnesia, however, is mainly to be found, with the 

 soda, in the ocean. On the whole the comparison is satisfactory, 

 although the lowering of the alumina is unexplained. The least 

 satisfactory feature of the combination is the average composition 

 of the shales ; but even here the percentage of soda is probably good. 

 In the important igneous rocks the proportion of soda varies within 

 narrow limits, and any combination of them would yield sedimentary 

 residues in which that oxide should show much the same relative 

 loss. The Survey average for soda (column £) gives 3.49 per cent. 

 The mean of 100 rhyolites given in Washington's tables yields 3.48 

 per cent., while for 220 basalts the percentage is 3.24. From the 

 preceding table the ratio between igneous and sedimentary soda is 

 1. 21 to 3.49, an order of magnitude which is not likely to be very 

 much changed. The influence of certain variations in it will be con- 

 sidered later. 



From the quantity of sodium in the ocean the mass and volume 

 of igneous rock needed to furnish it are easily computed, and also 

 the similar data for the sedimentaries. The oceanographic data are 

 fortunately quite good, even though they may not be rigorously 

 exact. The mean density of sea water, according to Murray, is 

 1.026, and its average proportion of saline matter is 3.5 per cent, by 



