92 CORRELATION OF GEOLOGICAL FAUNAS. [bull. silo. 
tion had not ceased, while the faunas above and below in the cliffs in Chemung 
Narrows is evidence that the geological horizon is that of the typical Chemung 
formation. The lapping of faunas of the same kind seems to be established by- 
evidence beyond dispute, and correlations must be made with recognition of such 
a possibility in cases where the direct evidence of the fact may be wanting. 
When we attempt to correlate formations with this knowledge before us it is 
evident that the life period of a fauna is not what it appears to be in any partic- 
ular section. Whenever the succession is sharply defined by the stopping of one 
fauna and the abrupt beginning of another, in full or decided strength, the evi- 
dence should be interpreted as positive that the boundary between the two con- 
secutive formations does not make the end of one fauna and the beginning of the 
succeeding one. It is to be interpreted rather as only a well-advanced stage into 
the later one and the vigorous period of persistence of the other. This, inter- 
preted into comparative terms, would result in showing that the two faunas lap 
over each other in time. 
My studies convince me that this is frequently the case in respect to the bound- 
ary lines of our formations. The abrupt transition from one formation to 
another with a different fauna is convincing evidence that the abruptness of the 
change in fossils is due either to absence of strata (i. e. . an apparent or concealed 
unconformity) or else to migration of the faunas across the area. 
This principle must be recognized in making correlation, if we would reach 
correct interpretation of the facts/' 
MARINE FAUNA ABOVE ONEONTA SANDSTONE OF EASTERN NEW 
YORK. 
Accepting Table XXI as an approximately correct list of the domi- 
nant species of the Spirifer clisjunctus fauna, as it existed in the 
typical area of its distribution, what relation does the fauna occurring 
above the Ithaca fauna in the eastern part of the State bear to it? 
In opening the discussion of this question it maybe noted that 
among the 20 dominant species listed in 'Fable XXI (the Spirifer 
disjunctus fauna), three are reported' by Gra ban from the Hamilton 
formation of Eighteenmile Creek. These are Ambocoelia umbonata, 
Chonetes sciiulus, and Palczoneilo constricta. The same species, and 
the variety arctistriatus of Orthothetes chemungensis are reported from 
the Hamilton faunules of the Cayuga Lake section by Cleland. All 
four of these species are specifically identified by Prosser in the 
Hamilton faunules of eastern New York and Pennsylvania. 
Removing from the list these recurrent species (viz., Ambocalia 
umbonata, Orthothetes chemurxje us is, Chonetes sciiulus, and PalcBoneilo 
constricta), as occurring also in the fauna of the Hamilton formation 
below, the remaining 16 will stand as characteristic species as well 
as dominant representatives of the typical fauna of the Chemting 
formation. 
In the sections in Chenango and Otsego counties above the Oneonta 
sandstone occasionally a few species occur which have led to classify- 
ing the beds holding them in the Chemung formation. 
In the recent revision of the geological mapping of that part of the 
State the State paleontologist appears to have adopted the Oneonta 
"Am. Jour. Sci., 4th series, Vol. XIII, 1902, pp. 428-431. 
