and the probable causes of their Variations. 209 
known. But in most cases, the re-union has been complete. 
Such remnants as the Glacial epoch left, have been equal to the 
emergency of perpetuating their races over the region desolated 
by glacial action, and they may thus indicate what are the 
possibilities of development under determinate conditions. It 
may be suggested, that as the species of so-called Strepomatide 
of the west coast have rather the facies of the tropical Melan- 
ians, and as the other associates of the J. margaritifera in the 
waters of Oregon are species not elsewhere found, that this 
little faunal remnant is an independent one, and I readily agree 
to all this; yet there is no doubt of the existence of a Fauna 
B, or of its distribution, and the possibility that its present 
species are the descendants of a geological remnant like those 
of A. Still more striking is the evidence to be adduced from 
Fauna C. The region over which this group is distributed 
tures, we should have another proof of the existence of what 
have been so philosophically called “comprehensive types ;” 
and it is by no means a difficult thing to show abundant 
evidences of their presence in this heterogeneous host of their 
modified descendants, as I hope to point out hereafter. Even 
if this fauna does not antedate the Carboniferous epoch, 
rope, new light will begin to break in upon the “origin of 
species” among these protean bivalves; for such work is the 
