4 
THE MIOCENE BEDS OP 
due to hot ferruginous water connected with the injection of 
the quartz-veins. 
The earth movements which gave rise to the deepening 
of the lake were doubtless responsible for the great activity 
of calcareous springs depositing frequent beds of travertine 
above the red clay (e.g. at Kachuku). Whenever the travertine 
became mingled with clay deposited at the same time, bands of 
hard, brown marlstone (Nos. 32 and 34) were the result, alter- 
nating with brown clay and enclosing shells of Ampullaria 
ovata , Lanistes carinatus and Cleopatra bulimoides. 1 Whilst 
these beds are very clayey at Nira they are represented at 
Kachuku by pebbly sandstones, showing that the river which 
brought down the sediments must have flowed from east to 
west, and in all probability it followed a very similar course 
to the present Kuja River. 
The most important beds of the whole series are the 
torrential and current-bedded sandstones and gravels of 
No. 31, which are particularly well displayed in the gullies of 
Kachuku (Fig. 4), for they comprise the zone in which I found 
bones of Dinotherium Hobleyi (mandible, tusk, &c.), and of 
Anthracotheres of different sizes, some allied to Hyopotamus 
(humerus, tibia, rib, and tusk) and probably leading a very 
similar existence to the present hippopotamus, and a small 
mandible of a form similar to Ancodus, a tooth of the hornless 
rhinoceros (Aceratherium) , the mandible of a small cat-like 
carnivore closely allied to Pseudaelurus, the astragalus of a 
Creodont, part of the carapace of a giant tortoise, scutes of 
Trionyx , teeth of crocodile, &c., and a very few landshells 
(Cerastus cf. Moellendorffi and Limicolaria), as well as the 
lacustrine Ampullaria ovata and Cleopatra bulimoides. 
The upper limit of these bone-bearing beds is readily 
recognisable, for it is formed by a thick conglomerate (No. 30) 
of white calcareous nodules (with concentric coats) from an 
inch or two up to 2 feet in diameter. 
The currents must have been particularly strong at this 
time to keep such large nodules in active motion, so as to 
permit the formation of this exceptionally coarse oolite in 
1 The vertebrate remains which I collected have been named by Dr. C. W. 
Andrews, and the mollusca by Mr. R. Bullen Newton. 
