FROM TANGANYIKA AND KENYA n 



Basement System by the Mansa Guda formation (Ayers 1952 : 6 ; Thompson & 

 Dodson i960 : 15), a series of sandstones and conglomerates, some 1300 feet in 

 thickness, which so far have yielded no fossils. The Mansa Guda formation may be 

 the equivalent of the Lugh Series of Stefanini, consisting of some 400 ft. of sand- 

 stones, marls and limestones developed to the east, in Somalia. Stefanini (1932) 

 recorded a small mussel-like bivalve and a naticiform gastropod from these beds 

 and thought that their age might be Lower Liassic. The fossils are not, however, 

 diagnostic and might equally well be of Triassic age. In the coastal area of southern 

 Kenya the Duruma formation, except for part or all of its top division, the Mazeras 

 Sandstones, is probably of much the same age as the Mansa Guda formation. In 

 Tanganyika the Jurassic rocks are underlain by beds of the Karroo System. There 

 is at present no fossil evidence that the Karroo beds extend above the Trias. 



The Lower Toarcian Didimtu Beds of N.E. Kenya were discovered by P. E. Kent 

 and F. M. Ayers in 1951 and first recorded by the latter (Ayers 1952 : 9). They 

 have been described in more detail by Thompson & Dodson (i960 : 20), who quote 

 (: 22) a preliminary report on the Bivalvia and Gastropoda by the present writer. 

 These fossils are described in the present work and listed on p. 189. Of the 30 

 named species now recorded from Didimtu, 22 are described as new and eight (one 

 with the qualification " aff. ") are referred to forms described previously, one of 

 which is re-named. Seven of these are also known from Europe. The eighth, 

 Weyla ambongoensis, a representative of the Pectinidae, was originally described 

 from Madagascar and is also found in Pakistan and Morocco. It affords somewhat 

 meagre evidence that a faunal sub-province comprising the western part of the 

 present Indian Ocean region and extending over northern Africa had come into 

 existence. Affinities with the Lias of Morocco are also indicated by the occurrence 

 of the new gastropod genus Africoconulus, the type-species of which occurs in the 

 Domerian of that country. The Didimtu fauna includes a rather larger assemblage 

 of Toarcian gastropods and bivalves than the contemporaneous fauna from Mada- 

 gascar described by Thevenin (1908ft), which consisted of 18 bivalves and two gastro- 

 pods. 



Bajocian Assemblages 



The Upper Bajocian age of beds included in the Kambe Series, developed in the 

 coastal district of Kenya, was established on the basis of ammonites collected by 

 Miss M. McKinnon Wood. The bivalves and gastropods from her collections, 

 amounting to 22 and two species respectively (some, however, identified only generic- 

 ally), were described by Weir (1930, 1938). No specimens from these beds have been 

 examined in the course of the present work. The Kambe Limestone is, however, 

 underlain by the Mazeras Sandstones, yielding fossil wood considered by its most 

 recent students to be Upper Triassic in age (Caswell 1956 : 16), although it was 

 thought that the upper limit of the Sandstones might lie within the Lower Jurassic 

 (Caswell 1953 : 17 ; 1956 : 17 ; Williams 1962 : 10). A sample of hard sandstone 

 belonging to this formation and found at the locality Ribe, about 9 miles N.E. of 

 Mazeras, has yielded a small series of gastropod moulds, one of which is described in 



