﻿270 DK. A. HOLMES ON THE TERTIARY [vol. lxxii, 



A felsie limit is reached at SiO.,=71'2, where the lime-curve meets 

 the base-line. The composition of the limit is that of a rhyolite. 

 Alumina is much higher and ferrous oxide much lower than in the 

 corresponding felsic limit of Series A. This contrast is in accord- 

 ance with the mineralogical characteristics of the rocks of Series A. 

 soda-pyroxenes and amphiboles being present, whereas in Series B 

 the ferromagnesian minerals are augite, hornblende, and biotite. 

 No attempt was made to find a mafic limit to Series B, for the 

 curves do not reach even the ordinate corresponding to 50 per cent, 

 of silica. 



Soda is again in excess of potash and the two curves are nearly 

 parallel, but converge as silica increases. The lime-curve is every- 

 where at a higher level than in the corresponding range of Series A. 

 The iron-oxide curves are similar in the two series, but that of 

 Series B is at a lower level throughout and declines more rapidly. 

 The alumina-curve of Series B is much flatter than that of 

 Series A, because in the latter the place of alumina is partly taken 

 by ferric oxide in the soda-amphiboles and pyroxenes. The general 

 form of the curves in Series B is very similar to those of the 

 volcanic rocks of Lassen Peak (California), 1 and of the San Fran- 

 cisco district of Arizona. 3 



Regarding each of the series as a whole, it seems doubtful — in 

 the absence of any knowledge of the relative volume of each rock 

 species — whether one can be regarded as more alkaline than the 

 other. The main distinction is that in Series A the antipathy of 

 lime and magnesia to the alkalies is much more marked than in 

 Series B, the separation of these two groups of constituents taking 

 place much sooner and much more completely in Series A than in 

 Series B. Each of the alkalies begins to exceed lime in Series A 

 at about 54 per cent, of silica, whereas in Series B this does not 

 occur until silica has reached 66 per cent. 



There seems to be no justification for supposing that the two 

 series of rocks are the result of different processes acting 

 on similar deep-seated materials, for they are both associated with 

 the same meridional system of faulting. We are, therefore, driven 

 to consider the two series as the consequence of similar processes 

 acting on different materials, those materials being in the 

 one case under-saturated with silica, relatively deficient in water, 3 

 and rich in carbon dioxide, and in the other case oyer- saturated 

 with silica, rich in water, and practically free from carbon dioxide. 

 Whether these two groups of parent materials were derived from 

 a pre-existing common source, is a problem that cannot yet be 

 definitely solved. 4 



1 A. Harker, ' Natural History of the Igneous Rocks ' 1909, p. 126. 



2 H. H. Robinson, U.S. Geol. Surv. Prof. Paper 76 (1913) p. 181. 



3 N. L. Bowen writes : ' It may well be that volatile substances are 



as abundant in magmas of this [alkali] series as in any others, but that water 

 is relatively an unimportant fraction ' Journ. Geol. Chicago, Suppl. to vol. xxiii 

 (1915) footnote 2, p. 78. 



4 For a suggestion that may apply here, see J. W. Evans, Geol. Mag- 

 dec. 6, vol. iii (1916) p. 189. 



