Abstract There are only a few areas on Earth where continental plate break-up and the attendant magmatism are taking place at the present day. In this context, the East African Rift system takes pride of place as it is the most extensive, presently active, continental extension zone, the extension being accompanied by seismicity, crustal thinning and, in some sectors, magmatism. The reason for this spectacular fracturing of the African Plate has been the subject of much debate but there is general consensus that it is due, at least in part, to the presence of rising thermal plumes in the mantle beneath Africa. The extension is now held to be due to the incipient separation caused by the plume-related eastward drift of the Somalia microplate away from the more stationary Nubian Plate (Fig. 1.1 ). The attendant fracturing extends from the Afar triple junction in the Red Sea in the north to at least the Zambezi River in the south. It splits into two branches around the Tanzania Craton which forms the topographic high of the East Africa Plateau, which itself is part of the larger highland area that covers much of southern Africa, referred to as the African Superswell ( Nyblade & Robinson 1994 ). The eastern branch of the fracture system, stretching from the Gulf of Aden, through Ethiopia and Kenya to northern Tanzania, mainly follows the north–south trend of the Mozambique Fold Belt and passes over the localized uplifts of the Ethiopia and Kenya Domes where the sub-parallel rift

Abstract During the Neogene–Recent there have been two major phases of crustal deformation in northern Tanzania. Each deformation event has been followed by a major pulse of volcanism arising from instabilities in the mantle. The chronology of the development of the rift faulting has been made possible by the dating of volcanic rocks that erupted before and after particular episodes of faulting. Appendix 1 ( Dawson 2008 ) lists the current dating results for the volcanic rocks. However, most of the dates were obtained many years ago by the K–Ar method and more refined stratigraphy will doubtless arise from Ar/Ar dating in the future. Whereas in Kenya, there is evidence for an elongate pre-rift depression in which the earliest volcanics were deposited ( Baker 1986 ), there is at present only sparse evidence for the presence of pre-volcanic sedimentary basins in northern Tanzania. The limited evidence comes from an exposure at the base of the rift escarpment at the north end of Lake Manyara, where a sedimentary formation, termed the Manyara Group, comprises a boulder conglomerate unconformably overlying metamorphic basement rocks. The conglomerate itself is overlain by siltstones, sandstones, waterlain tuffs and ashes with an intercalated basalt that has been dated at 4.86 ± 0.24 Ma. This is taken to indicate that, in this area, limited basin subsidence (? related to faulting) and basaltic volcanism had begun by around 4.9 Ma ( Foster 1997 ; Foster et al . 1997 ). This shallow-basin formation follows the oldest volcanism in the area, the eruption