﻿part 3] 



VOLCANIC ROCKS OF MOZAMBIQUE. 



271 



XX. Origin of the Lavas. 



Four problems, or groups of problems, present themselves for 

 discussion in this section : — 



(a) The differentiation of the alkali-lavas and their associates, Series A. 



(b) The origin of the parent magma of Series A. 



(c) The differentiation of the amygdaloidal basalts and their associates, 



Series B. 



(d) The origin of the parent magma of Series B. 



(a) The serial relationships of the lavas of Series A, disclosed 

 by their variation-diagram, point to some continuous process of 

 differentiation acting on a common parent magma. The sequence 

 of lavas in the Sanhuti district — (1) basalt, (2) phonolite, (3) 

 picrite-basalt — suggests that the first lava to be erupted most 

 nearly approaches the composition of the parent magma, and that, 

 the lavas that succeeded it stand in a complementary relation one 

 to the other, the one being rich in alkalies, the other in lime and 

 magnesia. The order of crystallization of the more significant, 

 minerals in the series regarded as a whole is approximately as 

 follows : olivine, titaniferous augite, labradorite, diopside, augite, 

 less calcic labradorite (andesine in tephritic pumice?), anorthoclase 

 and nepheline, soda-pyroxenes, soda-amphiboles, and opal (in the 

 ground-mass of aegirine-trachyte). It thus appears that magnesia 

 was the first of the main constituents to commence its withdrawal 

 from the parent magma, and that it was closely followed by lime. 

 A concentration of alkali alumino- silicates and of soda ferro- 

 silicates accompanied by increasing quantities of the volatile 

 fluxes thus became possible. Towards the close, water began to 

 exert a controlling influence, shown by the formation of soda- 

 amphiboles and still latei* by the deposition of natrolite. The 

 close genetic association of rocks like ajgirine-trachyte with free 

 silica (opal) and of phonolite with deficient silica (nepheline) may 

 also be due to the high-temperature activity of water. Describing 

 the history of a magma gradually modified by ciystallization, 

 Dr. N. L. Bowen 1 says 



' For any given concentration of the molecules KAlSi 3 O s and NaAlSi.,0 8 , 

 however small, there is a certain corresponding concentration of KAlSiO t , 

 NaAlSi0 4 , and Si<V 



the polysilicate molecules being broken down in the presence of 

 water. Later, he says (op. cit. p. 56) : 



' The precipitation of KAlSi.,0,,, NaAlSi :i O s , KAlSiO ( in mica, and Si(X as 

 quartz, means the concentration in the liquid of all the other molecules,' 



among which nepheline and the volatile fluxes are important 

 members. In preference to supposing that silica can be removed 

 from a magma by crystallization in the presence of water and other 



1 ' The Later Stages of the Evolution of the Igneous Rocks ' Journ. Geoh 

 Chicago, Suppl. to vol. xxiii (1915) p. 45. 



