OF DENISON UNIVERSITY. 
lOI 
are limited to this horizon or horizons still debated, 45. The division 
III, requires more study, but at least 7 species with carboniferous 
habit and 10 with no special bearing are unmixed with a species of 
Devonian alliances so far as known, except near the base. The evi- 
dence here is, it is true, chiefly negative. These conclusions have 
been reached on as conservative a basis as possible. If there is any 
error it is rather in favor of the carboniferous tincture in the two lower 
members. 
The tentative conclusions now held are, therefore, as fol- 
lows ; The Waverly group of Ohio is a composite assemblage of 
lithologically constant character. The lower portion of it is chiefly 
composed of greyish, yellowish, and greenish, arenaceous shales, with 
local grits and nodulary masses of limestone, and occasionally near 
the base intercalated layers of bituminous shale. This series is faun- 
ally nearly distinct in Central Ohio and should be regarded not only 
as Devonian, but as containing persistent elements of the Hamilton 
types in connection with Portage and Chemung forms. However, 
it is believed that geographical variation must be called in very 
largely to explain the specific divergencies, while generic resem. 
blances remain perfectly obvious. 
Our division H, although so relatively small, was evidently a transi- 
tion period. Most of the strata may have been deposited while the Cats- 
kill was forming at the east, but the fauna was essentially of Chemung 
character. Nevertheless the connection in Ohio was much more direct 
with areas where already carboniferous types were appearing and a more 
or less marked admixture was the result. Conglomerate H marks a 
slight oscillatory wave passing perhaps from north to south, resulting 
first in mud flats in which burrowing mollusks thrived, later in shore- 
lines kept supplied with pebbles by re-invigorated rivers. When 
the sea next returned it was with its freight of carboniferous forms, 
but the old Chemung species had chiefly perished during the slight os- 
cillation. Comparatively few deep-sea forms accommodated them_ 
selves to the littoral conditions for a time, but such as did are related 
with Burlington or Keokuk species, which formations were then accu- 
mulating to the westward. It is not difficult to conceive that condi- 
tions might exist in Ohio, which would considerably alter specifically, 
a fauna pretty closely concordant generically with the Chemung in 
New York. In our second division a number of species are specific- 
ally identical with those in New York. Yet we prefer to believe that 
