60 DR. A. HOLMES OlS" THE PRE-CAMBRTAlSr [vol. lxxiV, 



The constituents of these molecules were there, together with an 

 excess of silica, and potash was driven out, just as in the fusion of 

 a rock for alkali determinations by the Lawrence -Smith method 

 the alkalies are liberated from their alumino-silica union by inter- 

 action with calcium carbonate at a red heat. The following 

 equation indicates the kind of interaction that may have taken 

 place : — 



2(K 2 . A1,0 3 . 6SiO,) -f llCaC0 3 

 Constituents of Orthoclase Calcite 



= Ca 3 Al 2 (Si0 4 ) 3 + CaAl,(Si0 4 ) 2 + 7CaSi0 3 + 2K 2 C0 3 -4- 9C0 2 . 

 Grossularite Anorthite (for 



plagioclase) 



The figures in Table VI further suggest that, although the con- 

 stituents of albite have undergone a similar change, it must have 

 been to a less degree and also at a later stage. From Nos. 202 to 

 189 there is a considerable increase of the albite molecule, which 

 is probably more apparent than real : that is to say, the increase is 

 relative to quartz and orthoclase, and may not be an absolute 

 increase. From Nos. 1S9 to 194 there is a marked decrease in 

 albite, the change accompanying an increase of lime minerals. 

 Clearly the lime took the place of potash before soda was appre- 

 ciably displaced, and throughout the transition, sodic minerals do 

 not entirely disappear until the forsterite-marble itself is reached. 

 The selective preference of lime for soda is possibly related to the 

 fact that these constituents form isomorphous compounds in the 

 felspars and scapolites. There is, moreover, experimental proof 

 that the selection takes place, and that soda is retained by lime- 

 stone to a greater extent than potash. K. Endell 1 studied the 

 interaction between molten soda-microcline and limestone immersed 

 within the viscous mass. On cooling, the limestone was found to 

 have been replaced by a yellow glass, which, when analysed, proved 

 to be much richer in soda relatively to potash than the original 

 felspar. 



The changes accompanying the transition described above 

 clearly result in the liberation of solutions rich in potash, and, 

 to a less extent, in soda. Of the effects of these solutions there is 

 now no evidence ; but in other localities they may provide a clue 

 towards solving the difficult problem of the origin of highly - 

 potassic leucite-bearing lavas, such as those of the Bolsena- 

 Vesuvius district. 



In the Haliburton-Baneroft district of Ontario, F. D. Adams & 

 A. E. Barlow 2 have described the formation of amphibolite by the 

 interaction of granitic magma with the invaded Grenville Lime- 

 stone ; but there the relations are more complex, and the transi- 

 tions are not so regular as that described above. The} r say : 



' The crystalline limestone can be seen, under the influence of the granite 

 intrusion, to have changed into a typical hornblende-felspar amphibolite, 

 passing through the intervening stage of a pyroxene- scapolite-hornblende- 

 felspar amphibolite (pyroxene-scapolite-gneiss).' 



1 Neues Jahrb. vol. ii (1913) p. 152. 



2 Geol. Surv. Canada, Mem. No. 6, Pub. 1082 (1910) p. 104. 



