622 PALHONTOLOGICAL GEOLOGY. — [Boox V. 
compare the general order of its succession as a whole with that which. 
can be established for other provinces. It is, therefore, only after 
long and patient observation and comparison that the geological 
history of different quarters of the globe can be correlated. 
4, Subdivisions of the Geological Record by means © 
of fossils.—As fossil evidence furnishes a much more satisfactory 
and widely applicable means of subdividing the stratified rocksof the 
earth’s crust than mere lithological characters, it is made the basis 
of the geological classification of these rocks. Thus a particular 
stratum may be ascertained to be marked by the occurrence in it of 
various fossils, one or more of which may be distinctive, either from 
occurring in no other bed above and below, or from special abundance 
in that stratum. These species may therefore be used as a guide to 
the occurrence of the bed in question, which may be called by the 
name of the most abundant species. In this way a geological horizon 
or zone is marked off, and geologists thereafter recognize its exact 
position in the series of formations. But before such a generalization 
can be safely made, we must be sure that the species in question 
really never- does characterize any other platform. This evidently 
demands wide experience over an extended field of observation. The 
assertion that a particular species or genus occurs only on one 
horizon, or within certain limits, manifestly rests on negative evidence 
as much as on positive. The paleontologist who makes it cannot 
mean more than that he knows the species or genus to lie on that 
horizon or within those limits, and that, so far as his own experience 
and that of others goes, it has never been met with anywhere else. © 
But a single instance of the occurrence of the fossil on a different 
zone would greatly damage the value of his generalization, and a 
few such cases would demolish it altogether. ‘he genus Arethusina, 
for example, had long been known as a characteristic trilobite of the 
lower zones of the third or highest fauna of the Bohemian Silurian 
basin. So abundant is one species (A. Koninckz) that Barrande 
mentions that he had collected more than 6000 specimens of it, 
generally in good preservation. But no trace of it had ever been 
met with towards the upper limit of the Silurian fauna. Even- 
tually, however, a single specimen of a species so nearly identical 
as to be readily pronounced the same was disinterred from the 
upper Devonian rocks of Westphalia—a horizon separated from 
the upper limit of the genus in Bohemia by at least half of 
the vertical height of the Upper Silurian and by the whole of 
the lower and middle Devonian formations... Such an example 
teaches the danger of founding too much on negative data. To 
establish a geological horizon on limited fossil evidence, and then to 
assume the identity of all strata containing the same fossils, is to 
reason in a circle and to introduce utter confusion into our interpre- 
tation of the geological record. The first and fundamental point is 
to determine accurately the order of superposition of the strata. 
’ Barrande, “ Réapparition du genre Arethusina,” Prague, 1868. 
