1847.] SHARPE ON THE PALEGZOIC ROCKS OF N. AMERICA. 157 
Mr. Conrad * considers all the beds from the Cauda-zalli grit to 
the Tully limestone inclusive to correspond with our Ludlow forma- 
tion, leaving for the Devonian system the Portage and Chemung 
groups and some rocks in Pennsylvania called there Old red sand- 
stone. In favour of this view may be quoted one or two Ludlow 
species found in the Hamilton beds, and a large number of bivalves 
found in the same beds of the genera Avicula, Pterimzea, Cypricardia, 
Sanguinolaria, &c., which have a strong general resemblance to shells 
found in the Ludlow rocks of Westmoreland: but as very few of 
these species are identical with ours, and they are associated with 
many Devonian species, the balance of evidence is strongly in favour 
of classing the whole as Devonian. Moreover, so great a number of 
species are common to the Hamilton and Chemung groups, that they 
must both be classed in the same formation. 
It will be found safer to give up the attempt to identify any of 
the New York beds with our Ludlow formation, and to divide the 
“New York System” of rocks into three great divisions, equivalent 
to the Lower Silurian, Wenlock and Devonian systems of Europe ; 
the lines between which can only be drawn somewhat arbitrarily, 
owing to the gradual passage of all the groups into one another, and 
their general conformity throughout. ; 
It has just been stated that Mr. Lyell’s collection contains the 
following proportion of known European species of shells, viz.— 
The Lower Silurian series........ 30 per cent. 
siitie Wenlock Series: 0). <..+ 4~ « AO per cent. 
The Devonian series............ 20 per cent. ; 
but I wish it to be understood that I attach little importance to 
these or any similar numerical calculations of species. The sources 
of error in such calculations are numerous and varied, and so little 
known beforehand, that we can never be enough on our guard against 
them. Fossils must be collected to a far greater extent, and their 
comparison must be carried on with more care than has yet been 
given, before we can rely with safety on any calculations of the pro- 
portions of species. 
There are many circumstances which should make us mistrust the 
relative proportion of European species found in the three great divi- 
sions of the New York system. In this country the fossils of the 
Wenlock beds are easily found in good condition; they have in con- 
sequence found favour with collectors, and have been more thoroughly 
described than those of the other paleeozoic formations: they offer 
therefore more extended means of comparison than the Lower Silu- 
rian or Devonian species. The Lower Silurian fossils still undescribed 
are known to be very numerous ; but a large proportion of the Welsh 
specimens are in bad condition, and their determination is very diffi- 
cult. Large numbers of the Devonian species also have been found 
im a condition which does not admit of satisfactory comparison 
with foreign specimens. So that although my tables show in the 
Wenlock beds of New York a great excess of European species 
over either the Lower Silurian or the Devonian beds, it is by no 
* Journ. of Acad. of Nat. Sciences of Philadelphia, vol. viii. p. 232. 
