Williams.] SHIFTING OF FAUNAS. 113 
the precise relation of limestone sediments to the shores, for there are 
no terrigenous materials in the sediments. The limestone, when pure, 
does not necessarily indicate great distance from land erosion, and it 
may not indicate distance from actual shore. 
In the discussion of the Cuboides zone and its fauna a I adopted, 
as a working hypothesis, the view that limestone sedimentation con- 
stitutes a fifth class lying beyond the black shale end of the series. 
I think, in general, this is borne out by the facts; still it must be 
observed that limestones form near coasts, and, under favorable 
conditions, in water not deep. 
Where limestones continue to form for long periods, during which 
some oscillation is evident, the associated fragmental material is fine 
grained, and the passage from limestone into terrigenous deposits is 
generally, if not always, through fine-grained sediments to coarse; 
often black shales are among the transition beds. As a working 
hypothesis it would appear still to be safe to regard limestones as al 
least in the same class with black shales on a basis of relative distance 
from shore, and as a means of determining the direction of the shift- 
ing of the faunas. 
This particular order of distribution of the conditions of sedimen- 
tation in relation to distance from shore line may require modification 
as the facts are more thoroughly elaborated, but that the several con- 
temporaneous faunas associated with distinct types of sedimentation 
have shifted together laterally seems to be established beyond ques- 
tion. The following facts seem to favor this view: 
(1) Fossil faunas give indication of their normal association with 
particular classes of sediments. 
Unless we suppose that the fauna has shifted its local habitat the 
abrupt termination of a class of sediments in a given section requires 
the assumption that the fauna ceased to live, whereas, the actual 
continuity of life of species associated in faunal aggregates is theo- 
retically an established fact. 
(2) Sediments of each class are of limited geographical distribution. 
This fact taken with No. 1 makes the following a rational conclu- 
sion, viz: 
(3) A fauna in its purity is restricted in its geographical distribution . 
If a fauna in its purity has a limited geographical distribution, the 
recurrence of the same fauna in a continuous section, after the 
occupation of the region by an entirely distinct fauna, can be 
explained only on the assumption that the fauna moved away from 
the region during the interval of occupation by the latter. 
(4) Such recurrences of faunas are establish < I fa < /*, as shown on the 
previous pages of / // is < I ist : 1 1 ss ion. 
(5) A formation {when understood to be a continuous series of 
"Bull. Geol. Soc. Am., Vol. I, 189<), p. 481. 
Bull. 210— 03 8 
