250 Till American (J col 0(J l l<t . October, isai 
8. It is also coinpareil by means of representative forms, or tlirough 
genera anil families. 
1(1 and 8(/. Tliese comjiarisons are strengthened if two or more faunas 
in sequence are found to be systematically related to the faunas of a 
standard series. 
9. Two faunas or floras otherwise related are compared in age through, 
their relation to the present life of their localities. This method was ap- 
plied by Lyell to Tertiary rocks. 
10. Faunas are correlated by means of their relation to climatic 
episodes, taken in connection with station. For example, boreal shells- 
found in latitudes below their present range are referred to glacial time. 
In general the limitations to accurate correlation by biotic methods 
arise from the facts of geographic distribution. Correlations at short 
range are better than those at long range. 
Biotic correlation by means of fossils of different kinds may have 
different value. In general, the value of a species for the purposes of 
correlation is inversely as its range in time, and directly as its range ia 
space. The value of a biotic group depends (1), on the range of its 
species in time and space: (2), on the extent to which its representative* 
are preserved. 
Prof. K. von Zittel spoke in reference to the biotic methods and gave 
his opinion of the relative value of plants and animals for i)urposes of 
correlation. He regarded plants as relatively unimportant. Among, 
animals those which are marine, lacustrine and land animals may be 
distinguished. Of these classes marine iuvertel)rates are most value- 
able for purposes of correlation. The vertebrates change rapidlj^ but are 
frequently altogether wanting. For instance, no vertebrates occur in the 
Alpine l)eds corresponding in age to those w'hich contain the mammalian 
fauna of the Paris basin. In certain lacustrine deposits invertebrates 
may be absent, and in such cases the vertebrate fauna is the surest guide- 
Baron de Geer emphasized the importance of a mmierical comparison 
between different species. The actual counting of individuals in a given 
formation is of great value. 
Prof. 3Iarsh expressed his agreement in general with the conclusions 
communicated by Prof, von Zittel, but would give special weight to 
vertel)rate fossils. In the Mesozoic and Tertiary beds of the Kocky 
^lountains he had fovmd that the vertebrates offer the surest guide for 
correlation. This is in part because invertebrates are either wanting or 
are lacustrine. Prof. Marsh in 1877 named a sequence of horizons 
after the most characteristic vertebrate genus in each which is confined 
exclusively to it. He presented an outline of such classification brought 
down to date with a section to illustrate vertebrate life in America. 
AFTERNOON SESSION. 
Mr. C. D. Walcott spoke of tlie value of plants for purposes of geo- 
logic correlation. 
Prof. T. McK. Hughes spoke of the present and growing tendency 
toward a natural classification. The evidence is complex and iucludes a 
