From the introduction
"In his revision of the turtle family Pelomedusidae, Loveridge (1941) reduced the number of species of the genus Pelusios to the rather unexpectedly low figure of four, without any geographical races. The species thus recognized were: Pelusios adansonii (Schweigger), Pelusios gabonensis (A. Duméril), Pelusios subniger (Lacépède), and Pelusios sinuatus (A. Smith).
In 1954, L. Müller and Hellmich demonstrated that Pelusios niger (Duméril and Bibron) was a valid species and not a synonym of P. subniger. In 1956, studying Congolese and central African material, I revived another species merged with P. subniger by Loveridge, Pelusios castaneus (Schweigger) and also described two new species Pelusios nanus and Pelusios carinatus. Thus, 15 years after Loveridge's revision, eight species rather than four were firmly established. Unfortunately H. Wermuth and R. Mertens in their checklist of 1961 rejected these latter findings without a word of explanation. In 1956 I had also expressed the opinion that Pelusios bechuanicus FitzSimons might be valid as well; more recently, I have been able to confirm this with a beautiful adult specimen from southern Angola (Laurent,1964).
Since the validity of castaneus, nanus and carinatus was not admitted in a major publication on turtles I felt it necessary to present more convincing evidence, if possible. This investigation resulted, as a byproduct, in additional arguments for the validity of P. niger, and in clues to the geographical variation of P. castaneus of which several subspecies can be profitably recognized, and in fact I discovered eventually that still another overlooked species was hidden in this composite."