518 THE DISTRIBUTIOxV, ETC., OP ALKALINE ROCKS, 



The idea therefore suggests itself that, in this period, many- 

 masses of the primitive earth-crust and of the earliest sediments 

 which were largely chemical precipitates of an alkaline nature, 

 were brought so deep into the interior of the earth's crust that 

 they were refused, and in many cases chemically attacked 

 adjoining rocks, variously acid and basic, and formed what is 

 known as a foyaitic magma. 



This fusion and mixing are supposed to have taken place at a 

 depth many times greater than that at which the assimilation of 

 the sedimentary beds by the New England granites took place. 

 The mixing must have been perfect, leading to the transformation 

 of all the alkaline salts of sedimentary beds and primordial 

 oceanic precipitates to silicates. 



It will be convenient here to summarise some supports for this 

 view. 



(1) True igneous rocks always possess certain chemical con- 

 stituents present in their composition within certain limits. This 

 fact is utilised to distinguish between schists derived from 

 igneous rocks and those of sedimentary origin. A deficiency in 

 any one of these chemical constituents or an enormous excess of 

 one of them indicates that a rock has originated from a chemical 

 precipitate or from a sediment. The forces of erosion and of 

 redistribution of detritus by means of water always have the effect 

 of concentrating particular constituents in particular beds. Now 

 it is found that alkaline rocks universally, except when formed 

 by mixture of magmas (like many trachy-dolerites), have a great 

 deficiency of magnesia and lime, and they are specially enriched 

 in soda, potash and halogens. 



(2) If alkaline magmas originated purely by differentiation 

 processes why was there such a prodigious expulsion of them in 

 the Eocene 1 As differentiation has gone on at all times, and 

 under much the same physical conditions, we should be led to 

 expect in all volcanic epochs approximately the same sequence, 

 and alkaline rocks intermixed with calcic rocks in some definite 

 order of eruption .and in much the same quantitative proportion. 

 This would particularly be so if the planetesimal hypothesis of 



