226 =F. R. Cowper Reed—Fauna of the Bokkeveld Beds. 
amongst the trilobites Z/omalonotus and two special subgenera of 
Phacops; amongst the brachiopods Spzrzfer, and members of the genera 
variously termed Rhynchospira, Retzia, Rensseleria, and Trigeria, while 
Lepto. flabellites is the most abundant form of all. Chonetes is also 
represented by several distinct species, and is not an uncommon form. 
The most prominent Mollusca are Bellerophon (Bucaniella), Conularia, 
Paleoneilo, and Nuculites, the two latter containing at least 16 species. 
All other groups are very poorly represented. 
Further research may somewhat alter the proportions of the various 
groups and reveal new species and genera, but is scarcely likely to 
lead to a complete re-casting of our conclusions. 
Retations to Devontan Faunas oF OTHER REGIONS. 
(a) South America. 
The close relations of the Bokkeveld fauna to that of the Devonian beds 
of South America have been brought out in comparing the trilobites, 
brachiopods, and mollusca separately, and they have been remarked by 
all workers on the fossils. Katzer (15, p. 28) has especially compared 
the Amazonian fauna with it, and concludes that they are of the same 
age. The intimate nature of the relations between the several 
Devonian faunas of South America itself has been emphasised by the 
same author as well as by Clarke (8a, pp. 80-90) and Thomas (10). 
(6) North America. 
The same authors lay stress on the resemblance of these various 
Devonian faunas of South America to that of the Hamilton Beds 
of North America. Kayser (20, p. 317) had previously insisted on this 
point, and we have seen the strength of the same resemblance in the 
case of the South African fauna. Moreover, the same peculiar admixture 
of Lower Devonian forms with those of Middle Devonian age is likewise 
noticeable in South Africa, but the latter predominate and give the 
principal features to the fauna of the Bokkeveld Beds. 
In South Africa there are species allied to or identical with those of 
the Upper Helderberg Group (including the Oriskany Sandstone), 
of the Hamilton Group, and of the Chemung Group; and Katzer 
(15, pp. 20-22) has recognised the same in the Amazonian faunas. 
The conclusion to which Thomas (10) is led with regard to the age of 
all the South American Devonian beds is that it is close upon the 
borders of the Lower and Middle Devonian. But Katzer is of the 
opinion that the Devonian fauna of the Amazon, and especially the 
Maecurt fauna, is more closely connected with the Middle than with 
the Lower Devonian of North America; and it is to this conclusion 
I am brought by the foregoing study of the affinities of the Bokkeveld 
fauna. The evidence of each zoological group, so far as it is known, 
points in this direction, and I believe it is largely the lithological 
resemblance of the beds to the well-known Spirifer Sandstone of the 
Rhenish Lower Devonian which at first inclines one to put the South 
African beds on the lower stratigraphical horizon. But to this feature 
reference is again made below. 
