28 N. L. Boiven — Genetic Features of 



formed by accumulation of their crystals from much more 

 salic liquids (nephelite syenites, etc.). The remarkable 

 development of flow structure in these rocks, which is 

 summarized for the whole province by O'Neill, 20 might 

 therefore be regarded as a natural result of their mode 

 of origin. If, as is suggested, they do not come into 

 existence except as a result of a considerable degree of 

 crystal accumulation of the appropriate kind, the devel- 

 opment of flow structure, when such a mass is moved 

 is inevitable. 



Formation of Lime-rich Minerals in Alkalic Rocks. 



The series of figures (1, 3, 4, and 5) serves to emphasize 

 the fact that in the presence of considerable nephelite 

 the very basic silicates forsterite, melilite and possibly 

 monticellite separate from liquids well over on the silica 

 side of the nephelite-diopside join. The presence in a 

 rock of the lime-rich mineral melilite need not, therefore, 

 be considered as indicating the existence of a magma par- 

 ticularly rich in lime but rather of an alkali-rich magma 

 from which melilite-bearing facies might be formed by 

 differentiation (crystal accumulation). Moreover, the 

 intimate association of melilite and nephelite rocks in 

 nature is to be regarded as further evidence of the impor- 

 tance of crystallization-differentiation, melilite being 

 a normal cafemic constituent of nephelite-bearing rocks. 



Even such minerals as garnet and vesuvianite, closely 

 related as they are to melilite, 21 may, perhaps, be capable 

 of formation as normal constituents from alkalic magmas 

 and are not necessarily an indication of contamination of 

 the magma by lime-rich rocks such as limestone. These 

 very basic silicates may conveniently be regarded as the 

 normal basic minerals of rocks (mainly metasilicates) 

 desilicated in the presence of alkalic molecules such as 

 nephelite. It will be noted that this action, which has 

 been demonstrated for the artificial mixtures, is exactly 

 the opposite of that which Daly supposes to take place in 

 magmas, viz., the desilication of feldspathic molecules by 

 lime with formation of feldspathoids. This fact should 

 not, however, in itself, be urged as necessarily disproving 



20 Geol. Surv. Can. Memoir, 43, p. 22, 1914. 



21 The sarcolite molecule of melilite is also a garnet molecule. 



