206 Proceedings of the Royal Physical Society, 



reason to l3elieve is really the case. No organic agency 

 yet known to us is engaged upon the work; and we are 

 therefore led to speculate whether the cycle of change 

 may not be completed by the gradual transfusion by osmosis, 

 of the superabundant alkaline matter, through the floor 

 of the ocean, most especially within the zone of chief 

 terrestrial mobility above referred to as the Transitional 

 Zone. Osmosis may thus be supposed to carry these aqueous 

 solutions downwards through the outer portions of the 

 submarine parts of the lithosphere, to those horizons beneath 

 the surface where the temperature sufficient for dissolving 

 rock-material is to be found. This depth need not be very 

 great, nor need the temperature be very high, for the 

 experiments made by Daubree, Delesse, and others, show 

 that alkaline solutions are quite able to dissolve silica and 

 most of the silicates — if sufficient time be allowed — at a 

 temperature some hundreds of degrees below that required 

 for dry fusion. (It is at least conceivable that the tempera- 

 ture exhibited by lavas when they reach the surface may 

 be far above that at which their liquefaction commenced 

 within the lithosphere : but upon this point, again, we have 

 little or no exact information.) It appears to me that nothing 

 more is required than such conditions as these, and I can 

 see no difficulty in accounting for the conversion of sediments 

 into gneisses, for their replacement by any kind of eruptive 

 rock, or for their conversion into those aqueous solutions 

 which give rise to lavas, if the factors just referred to 

 operate in the manner suggested. 



This view of the purely aqueous origin of the magmas of 

 eruptive rocks is alike compatible with the fact that, under 

 one set of conditions, fluidity may be generated at a lower 

 level, whence the magma is bodily transferred to a higher 

 by the explosive force exerted by its heated H^O, and, under 

 another set of conditions, what eventually becomes a crystal- 

 line eruptive rock may be generated and consolidated with- 

 out any such transference. In the case of the formation of 

 a sill or a dyke, for example, all that appears to be necessary 

 is the forcible injection of heated alkaline waters, in the first 

 instance, along some divisional plane of the rock that is being 



