12 CORRELATION OF GEOLOGICAL FAUNAS. [bum. 210. 
ually in various places and in different ways. The same experience is found, to a 
greater or less extent, as any local terms are extended from any of the States into 
those contiguous. This plainly shows that unless there he allowed great freedom 
to vary from the scheme adopted for stratigraphic designations, any nomenclature 
which the committee or the International Congress may adopt will he but a short- 
lived experiment. 
It will obviate all this confusion if * * * one set of names be chosen for the 
lithological characters and another for the faunal. 
The stratigraphic terms should be wholly geographic and should be allowed to 
change as often as local geologists deem it is necessary. The faunal terms should 
be very broad in their scope at the outset, and subdivisions should be introduced 
as fast as the special subfaunas are discovered and defined." 
This was stated more explicitly in a paper published in 1894.* 
As surveys have advanced, and as the field of geological correlation 
has gone beyond local and national boundaries, the task of establish- 
ing correlations has made the necessity of a dual nomenclature 
more imperative. Correlations between widely separated regions are 
now established on the basis of fossils alone. Correlations on the 
basis of continuity of lithological peculiarities are already known to be 
valid for only limited areas. Thus geologists throughout the world 
are already adopting the principle of a dual method of correlation, 
although the nomenclature and classification of correlation are still 
primarily conjoined with lithological formations, the names of which 
furnish the only means of distinguishing the faunas and floras which 
they contain. 
This lack of a nomenclature by which to distinguish the lithologic- 
ally defined formation from the biologically defined fauna (which 
may or may not be limited in its range by the boundaries of the for- 
mation) can be supplied only through discrimination of the charac- 
teristics of actual fossil faunas and a demonstration of their 
independence of the limiting conditions by which the formations are 
defined. If it can be shown that fossil faunas and floras can be dis- 
criminated, defined, and discussed separately from the formations, 
which now constitute the only elements of geological classification, 
not only will the separate nomenclature naturally follow, but the fos- 
sil fauna will then become, as it is now partially recognized to be, the 
definite means of determining the time relations of geological for- 
mations. Such a discrimination is attempted in the following pages. 
In order to exhibit the characteristics of faunas a concrete case is 
selected from among the faunas of the Devonian system, the choice 
having been determined by the abundance of the facts already gath- 
ered regarding Devonian faunas. Abundance of fossils, frequency of 
exposures, and wideness of distribution distinguish the Hamilton 
formation of the New York section above all other formations in the 
country. The large number of workers, the degree of refinement in 
analysis, and the fullness of publication of the statistics regarding 
aCompte Rendu Congres Geologique International, fourth session, 1888, A 91. 
bOn dual nomenclature in geological classification, by H. S. Williams: Jour. G-eol., Vol. II, p. 145. 
