In a recent memoir (Holmes and Harwood, 1937, p. 276) I had occasion to point out that “one of the most remarkable features of the igneous activity of late Cretaceous and early Tertiary times is the geochemical and petrological contrast between the kimberlite and olivine-melilitite which perforated the crust of Southern Africa and the widespread outpourings of olivine-basalt and basalt (often followed by acid rocks) in many other parts of the world”. An essentially similar contrast is found in the more recent and more localized volcanic fields of the Western Rift of Africa. The Toro-Ankole fields of multiple vents, east and south-east of Ruwenzori, are characterized throughout by undersaturated potassic types, free from felspar; whereas in the great lava fields south of Lake Kivu the dominant lavas are olivine-basalts of various types. Between these two strongly contrasted provinces, and in some respects sharing the characters of each, the Birunga field (including Bufumbira) extends across the Rift north of Lake Kivu. Although many of the lavas of this field contain abundant plagioclase and look like basalts, they have been found, with few exceptions, to be potassic types ranging from leucite-basanites to basanitoid trachybasalts. True basalts appear to be absent (Holmes and Harwood, 1937, pp. 16–18), the nearest approach to basalt—geographically as well as petrographically—being certain trachybasalts and limburgites from the south of the field, which differ from the other rocks of Birunga in having more soda than potash.