2298 ae Cowper Reed—Fauna of the Bokkeveld Beds. 
ica nee Cucullella elliptica, Maur. Lr. Dev. 
NV. capensis % 508 mais LV. solenoides (Goldf.). Lr. Dev. 
Buchiola subpalmata 306 340 B. palmata (Goldf.). Mid. Dev. 
Sanguinolites niger : aE S. Marsi, Oehl. Lr. Dev. 
Actinopteria aff. Boydi.. sae Act. Trigeri, Oehl. Lr. Dey. 
'. We have therefore to acknowledge a certain degree of affinity to the 
fauna of the Coblenzian Beds, but it is of much less strength than 
that shown to the North American beds. 
The problem, however, now confronts us of the alliance of the 
Bokkeveld fauna being closer with the Middle Devonian of North America 
than with the Lower, while in Europe it is with the undoubted Lower 
Devonian fauna that we find most points of resemblance. The same 
difficulty has to be faced with regard to the South American fauna, 
and Katzer (15) inclines to the explanation that the European Lower 
Devonian forms migrated to America, appearing and living on later 
there so as to form an element in the Middle Devonian of that area. 
Probably the physical conditions which they found, especially in South 
America, and we may add South Africa, were conducive to their 
persistence, as these resembled the Coblenzian of Western Europe. By 
this attitude Katzer is in favour of the non-contemporaneous appearance 
of the same or closely allied forms in Europe and America; and he 
suggests that the whole movement and dispersion of the faunas was 
connected with that widespread transgression of the Middle Devonian 
seas of which fresh evidence is constantly being found (25). It was, 
however, the typical West European Middle Devonian fauna which 
spread over the widest area, but its extension is found to the east of 
its well-known habitat in Western Europe, and not to the south or 
west. It has been discovered throughout Central Asia northwards 
to Shantung and southwards to Burma, and as far as Australia, 
a growing admixture of American forms being generally noticeable as 
we proceed eastwards. The characteristic calcareous development of 
this type is in marked contrast to the arenaceous or argillaceous nature 
of the South African and South American beds, and the physical 
conditions must naturally have largely influenced the composition of 
the faunas. 
The whole question of the migration of faunas, particularly im 
connection with the Devonian beds of North America, has been 
recently discussed in much detail by H. 8. Williams (31), and he has. 
shown by a wide collection of facts and statistics that there was 
frequent lateral shifting, emigration, recurrence, and overlapping of 
faunas living in the same general marine province, with occasional 
invasions of strange faunas from neighbouring or distant areas. 
Migrations have been usually attended by modifications of some of 
the species, or of their proportionate abundance ; other forms which 
were unable to shift or adapt themselves to new conditions with 
sufficient rapidity become extinct, while others are stimulated to 
mutation (in De Vries’ sense) and new types are produced. An 
incorporation of elements previously existing in the invaded area may 
also be occasionally noticed. In this way the survival of an older 
fauna in its integrity, or at any rate of a considerable remnant, may 
take place owing to lateral shifting, and the lifetime (biochron) of 
any fauna does not necessarily, or even generally, coincide with the 
