Vol. 70.] MIOCENE OF THE VICTOBIA NTA1TZA, ETC. 135 



time of a calm lagoon, communicating with the main body of the 

 lake to a sufficient extent to allow the molluscs AmpuUaria and 

 Lanistes to live in its waters and to be preserved in this bed. 

 The remark may be made, in passing, that no bivalves were pre- 

 served in any of the deposits — a circumstance difficult to explain, 

 in view of the abundance of Unio, ^Etheria, Splicer i um, etc., on 

 the shores of the present lake, unless the sediments were laid down 

 at such a distance from land that the only shells capable of being- 

 deposited were those of gastropods which floated until they became 

 waterlogged and eventually sank. At the present day, gastropod 

 shells are seen floating on the lake several miles away from the 

 shore. 



The red coloration of the clays is the most marked feature of the 

 middle series, especially in the western area, and perhaps indicates 

 the activity of ferruginous springs in this district. 



Finally, the upper series indicates the tune when the rivers 

 had nearly reached their base-levels, and were no longer able to 

 brins^ down heavv loads : the sediments consist essentially of o^ey 

 or brown clays alternating with shales, and interrupted quite ex- 

 ceptionally by grey sandstones, or more frequently by thin seams 

 of travertine. These sandstones with large plates of biotite and 

 augite-erystals probably were the product of unusually Avet seasons, 

 when the rivers were swollen beyond their ordinary dimensions. 

 In the uppermost beds the petrified stems represent some unwonted 

 circumstances, in which waterlogged tree-trunks were calcified by 

 the agency of calcareous springs. 



The upper series (averaging 70 feet) is nearly equal in thickness 

 to the middle and lower series taken together. This circumstance, 

 in conjunction with the argillaceous nature of the sediments, 

 indicates that a very long period of time elapsed during their 

 deposition ; therefore, although the lowest beds are of Lower 

 Miocene age, it is not impossible that the uppermost beds may 

 extend even into the Pliocene. 



III. DlSTEIBTJTIOX OP THE MlOCEXE SERIES. 



These Miocene deposits have been almost completely overwhelmed 

 by flows of nepheline-basalt, which do not, however, seem to have 

 appreciably baked the underlying clays for more than a few inches ; 

 but the junction was everywhere obscured by thick downwash and 

 overgrowth half way up the cliffs of East Kachuku and Kikongo. 

 The insignificant amount of baking was probably due to the fact 

 that the lava-streams had reached their southernmost limit, and 

 had therefore already cooled down considerably. 



The Miocene deposits are present along the southern margin of 

 the lava -plateau, from Nira on the lake to Kikongo, 41 miles away 

 to the east (PL XXVI, fig. 1), but unfortunately they are still for 

 the greater part concealed beneath a thick covering of superficial 

 deposit termed regur or black cotton-soil; and it is only in the 

 gullies of Nira, Kachuku, and Kikongo that a relatively small 



Q. J. G. S. No. 278. L 



