550 SCIENCE PROGRESS 



phases. The magmatic circulation that, with the aid of simple 

 cooling, is thus started, is believed to explain such phenomena 

 as the alternation of varying mineralogical bands, and the 

 parallelism of these bands to the walls of the magmatic chamber. 



In an interesting study of the origin of serpentine (Amer. 

 Journ. Set. 191 8, 46, 693-731) W. N. Benson emphasises the 

 view that the larger antigorite and chrysotile masses are due 

 to the alteration of pyroxenic peridotites through the agency 

 of magmatic waters appertaining to the same cycle of igneous 

 activity as the ultrabasic rock itself. In regard to the for- 

 mation of " serpentine " from olivine in volcanic rocks by 

 ordinary weathering, he doubts whether antigorite or chrysotile 

 varieties are formed in this way, although the minerals iddings- 

 ite and bowlingite, often recorded as serpentine, unquestion- 

 ably are. 



A. L. Du Toit describes an unusual occurrence of corundum- 

 aplite (plumasite) from Natal [Trans. Geol. Soc. South Africa, 

 191 8, 21, 53-73)- These rocks are dykes intrusive into ser- 

 pentine and granulitic gneiss. In contact with the latter rock 

 the dykes consist of normal quartz-bearing granite-aplite ; but 

 where they cut the serpentine quartz disappears, the felspar 

 becomes an acid plagioclase, and corundum appears in con- 

 siderable quantity. '1 his change is ascribed to the desilicating 

 influence of the serpentine upon the acid magma of the dykes, 

 an action which is confirmed by the transformation of the 

 adjacent serpentine to phlogopite and talc, requiring the 

 transference of large amounts of silica, alumina, potash and 

 fluorine, from the aplite magma. Some titaniferous magnetite 

 rocks are described in the same paper. Ihese are due to 

 magmatic segregation within a gabbro-pyroxenite mass on the 

 River Tugela. Green spinel, and the ferruginous olivine hyalo- 

 siderite, are included within the iron ores. P. A. Wagner 

 describes an occurrence of corundum in a plumasite-pegmatite 

 from the Zoutpansberg corundum fields of the Transvaal 

 {ibid. 37-42). This is a coarse rock containing andesine and 

 biotite beside corundum, and is mineralogically similar to 

 Lawson's plumasite from California, as well as to the rock 

 described in the preceding paragraph. 1 he field relations of 

 this rock are unknown, and consequently its mode of origin 

 cannot be determined. 



In a third collection from South Georgia the author has 



