I8.J4 VARIATION OF IGNEOUS ROCKS. 139 



I claim, also, that the same paste has acquired those elements that 

 have disappeared from the encasing rock. The change in composition 

 of a lava by one only of these processes would not be very great, 

 but when both are considered the change of an acid to a basic rock, 

 or vice-versa is quite possible. 



In my paper on the ejected blocks of Monte Somma^ (I. stratified 

 limestones) and that on eozoonal structure in the same, by Dr. J. W. 

 Gregory and myself, now being published by the Royal Dubhn 

 Society, it has been shown that calcareous rock acquires silica and 

 later alkalies, and loses lime and magnesia in such a way that the 

 magma, actually cooling in contact with it, is often reduced to an 

 ultrabasic one. Are we not to suppose, therefore, that a less intense 

 change does take place at a greater distance and under less favourable 

 circumstances than these extreme and rather localised ones ? Again, 

 when we examine great areas of intrusive acid rocks, we often find 

 their peripheral regions becoming more and more basic, and, perhaps, 

 cases may be recorded of the reverse, according to the country rock. 

 That these gradations are irregular is nothing more than we should 

 expect, for the country rock surrounding such a mass will also vary 

 in composition, and then circulation of the paste towards volcanic vents 

 may have been greater at some spots causing mixing or carrying away 

 the material. For succinctness we may call this the "osmotic" 

 hypothesis. 



If we suppose a number of vents, scattered over such an area, to 

 open, the outpouring lava would differ according as it was derived 

 from the purer or from the altered paste. A vent might drain out all 

 one kind of paste, and following this might flow another of a different 

 composition. The order in the composition of the rocks that issued 

 might depend, therefore, on quite accidental circumstances. That 

 not uncommon order of some regions commencing by the emission 

 of basic rocks that become more and more acid and then, after an 

 interval, terminate by a final phase of very basic rock, is quite 

 explicable on this hypothesis. Let us imagine a great laccolite of an 

 acid rock taking up its position among limestones. During a 

 considerable period interaction takes place and a peripheral layer 

 of basic rock is formed, being naturally basic at the surface near the 

 limestone. This outer layer also becomes more and more aquiferous by 

 the assimilation of H.^O from the enclosing rocks, until its tension 

 rises enough for it to burst its way to the surface. The first out- 

 pour will be eminently basic, but as this drains off, the more and more 

 central and acid layers will rise and follow. When sufficient has 

 escaped to reduce the tension of the whole laccolite, and this becomes so 

 viscous as to be unable to issue, the volcano will become quiescent, 

 during which basification of the acid rock and absorption of H^O will 

 still progress, rendering it more fusible, and then, becoming more fluid 



** Proc. Geol. Soc, 1888, pp. 94-96, and Trans. Edinb. Geol. Soc , vol. vi., p. 314, 1893. 



