Editorial Coin me nt. 253 
Neglecting this principle has led to the greater part of the mistakes 
■of paleobotanists, and has done most to bring paleobotanj' into disre- 
pute. Geologists have expected too much of them, and they, in turn, 
have done violence to the truth in attempting to satisfy extravagant de- 
mands. On the other hand, where the material is ample, fossil plants 
have often corrected the mistakes of stratigraphical geologists, and 
solved problems concerning geologic age, which seemed impossible of 
settlement by any other class of evidence. 
Mr. Chas. D. Walcott spoke upon the correlation of the Cambrian 
rocks of North America. The principles used now are the same as 
those used by the New York survey prior to 1847 except that those 
principles have been somewhat modified by the theory of evolution. 
Both physical and biotic data are available in the correlation of the 
Cambrian rocks on the Atlantic coast, of the Rocky ^Mountain areas, and 
of the interior continental plain. Throughout the Appalachian prov- 
inces the physical data suffice to correlate the Lower Cambrian from 
Vermont to Alabama, but such data are not sufficient to correlate it with 
that North of the St. Lawrence valley. The correlation of the deposits 
of the Appalachian and the Rockj' Mountains troughs were by biotic 
data alone, and of the great extent of the Upper Cambrian over the 
continent the biotic data correlated the Rocky Mountain Upper Cam- 
brian with that of the interior and the Appalachian area. 
The correlations made indicate that in Lower and Middle Cambrian 
time a 'great continental area existed over the interior, and all the Cam- 
brian sediments were accumulated in troughs west of the Appalachian 
and Rocky Mountains. In Upper Cambrian time the interior of the 
continent sank beneath the ocean, and the sandstones of the upper Cam- 
brian were deposited, and the result of these correlations add a chapter 
to the history of the evolution of the North American continent. 
Prof. James Hall spoke of the difficulties encountered in the earliest 
attempts at correlation of the rocks even in the state of New York. 
He urged the importance of taking into consideration both physical and 
faunal characters of the rocks. In some cases, however, the physi(;al 
characters of the rocks change greatly in passing from one region to 
another- -sandstones grading into limestones, and limestones into shales 
— and these beds may also vary greatly in thickness. Fossils are of 
unequal value in such correlations; Lamellibranchs are near shore forms 
and fail in deep water; they are not, therefore, so valuable as Brachio- 
pods, which have a wider distribution, for purposes of correlation. 
Prof. Henry S. Williams laid stress upon the relations of species to 
the conditions of deposition. The abundance of a species varies with 
environment, and a study of correlation should enil)race a study of 
these conditions. Sandstones deposited near shore may have a fauna 
different from that of a limestone deposited off shore at the same time, 
and a change of fauna may be induced by a change of the conditions of 
deposition. The age of beds should be determined by comparing 
species of the same genera rather than b)' comparing those of different 
genera. There are centres of abundance which exhibit great variability 
