THE TANGANYIKA PROBLEM. 
335 
was, at any rate once, wide-spread in Africa ; and that a 
large percentage at least of the halolimnic shells correspond 
to those in the upper cretaceous fresh-water beds. 
Dr. Gregory was consequently of opinion, before the 
second Tanganyika expedition started, that we should 
encounter the halolimnic fauna in other African lakes ; 
or at least in the old lake deposits which occur in asso- 
ciation with them, outside the region of Tanganyika 
altogether. This view, however, has entirely broken 
down, as it is now known we did not encounter the halo- 
limnic fauna outside Tanganyika. It is not present in 
Nyassa, Shirwa, or even Kivu ; it does not occur in Bang- 
weolo or Mwero. It is not present in the Albert Edward, 
the Albert or the Victoria Nyanzas. It does not occur 
in Beringo, nor yet in Rudolf. Neither did we find the 
remains of the halolimnic forms in any of the numerous 
old lake deposits which we examined or heard of, through- 
out these regions. 
There is thus, at the present time, not only evidence to 
show that the fresh-water fauna of the type of that occurring 
in the upper cretaceous beds is not now characteristic of the 
great African lakes, but, also, much to show that it never 
has been so at anytime. Further, the comparison between 
the halolimnic shells and those of the Southern European 
and American fresh-water cretaceous beds rests, as I have 
said, upon the flimsiest ground. Solely, as a matter of fact, 
on the supposed identity between shells of the Parame- 
lania of Tanganyika and the Pyrgulifera of these creta- 
ceous beds. 
I carefully examined all the figures of these shells, and all 
the fossil specimens contained in the collections of the 
British Museum from these districts. But, beyond the 
Pyrgulifera , could find no other shells among the cretaceous 
