Williams.] SHIFTING OF FAUNAS. 101 
but so far as such work lias already been carried the evidence is all 
against the supposition that the sandstone of the Otsego seel ion would 
be a sandstone in the Allegany County section. This is borne out 
in the special case of the Oneonta, which is lost as a red sandstone 
mass before reaching Tompkins County. 
We are therefore forced, by the evidence before us, to conclude 
that litliological characters, which constitute the basis of discrimination 
of the geological formations as units, not only can not be relied upon to 
discriminate time equivalency, but uniformity of litliological constitu- 
tion must be regarded, in some cases at least, as positive evidence of non- 
equivalency in time. This rule is applicable whenever the formation 
is traced at right angles to the original shore line along which the 
sediments were deposited. The exception to the working of the rule 
is in those cases where the formation is traced in a line parallel to the 
original shore line. In such a case sedimentation may have been 
approximately uniform for long distances. 
It is necessary, therefore, not only to use the fossils as an aid to 
stratigraphy in determining equivalency, but the fossil evidence must 
be so separated from inferences drawn from formation names that its 
real value in time discrimination can be independently estimated. 
To make such separation of the two sources of evidence of time 
relations (viz: formations and faunas), it is necessary to deal with the 
fauna independently of its particular place in any geological column 
of formations. So considered the fauna is an aggregate of organisms 
combined in such number of genera, species, and individuals as to 
express the bionic values of each in their relations to the total corpo- 
rate fauna of each epoch of time for the area covered. 
The presence of a few species which are common in the tj^pical 
series of rocks called the Hamilton formation (as currently defined by 
geologists) is not evidence of contemporaneity of formation for the 
rocks containing them in some other region. In fact, we have shown 
that the 12 most dominant and characteristic species of the formation 
actually do all occur in the Ithaca formation, which, at Ithaca, is 
separated from the Hamilton formation by two well-defined geological 
formations (the Tully limestone and the Genesee shale) and by still 
another series of rocks with a distinct fauna (the "lower Portage" 
so-called, with the Spirifer lmvis fauna) — in all about 400 feet 
of strata. Nor does the mingling of species of one fauna with those 
of another invalidate the value of the faunas as time indicators. 
Again, in order to use the fauna as a time indicator, the changes in 
the fauna coincident with passage of time must be observed and noted. 
The study of the details of these Devonian faunas, as has been already 
stated, brought out the fact that a fauna may retain for a consider- 
able thickness of sediments its integrity as a general fauna— viz, 
its corporate integrity. Illustrations are given in Grabau's and 
Cleland's analyses of the successive faunules of the Hamilton for- 
mation. In such range of a fauna through hundreds of feet of 
