386 H. S. WILLIAMS CORRELATION PROBLEMS 



particular combination found in the Edmunds, with less evidence of the 

 Wenldck forms peculiar in the Rochester f anna of New York. 



SUMMARY OF THE FACTS OF CORRELATION 



,n •>...• ' ".""■■ 



I have now presented enough of the facts to make clear the problems 

 to which I wish to call attention. 



The faunas of the Rochester r shale of New York and the; Edmunds 

 formation of Maine differ considerably in composition, but; a close com- 

 parison: of the 1 particular species of each shows that each of them finds 

 its. nearest correlate in the British Silurian at the same horizon, i namely, 

 the Wenlock limestone and immediately underlying shales. 



Examination of the species /distinguishing these two faunas show's that 

 the peculiar species of the Rochester fauna have closer affinity in the: 

 English section with the faunas below the Wenlock, and the peculiar 

 species of the Edmunds show closer affinity with faunas which follow the 

 Wenlock. 



The hypothesis suggested by the facts is that there are here concerned 

 two faunas which are fonnd on this side the Atlantic in separate areas of 

 distribution, but in England are blended together- in the typical Wenlock 

 district. • . 



This hypothesis seems to be supported.by the increasing dominance of 

 the Edmunds group of species in the Gotland district and the conspicu- 

 ous decrease in importance of the Rochester group of species still farther 

 east and south in Podolia ; also by the very slight expression of the Ed- 

 munds group of. species anywhere on the North American continent. 



It will be noticed that the correlation of the Edmunds and Rochester 

 faunas with the transatlantic Wenlock is established on statistical evi- 

 dence of identical species. 



The hypothesis of the distinctness of the Edmunds and Rochester 

 faunas is based on the statistics of range and distribution of the species, 

 namely: 



Certain species common to the Edmunds, Rochester, and transatlantic 

 Wenlock faunas are notably of long range below and above the Wenlock 

 and of wide distribution. 



The transatlantic Wenlock species found in the Rochester and not in 

 the Edmunds fauna are conspicuous for ranging lower in the series 

 rather than higher, for holding on in the northern extension of the 

 Wenlock of Great Britain, and for becoming less conspicuous in the far 

 east European distribution. On the other hand, the peculiar species 

 common to the Edmunds and Wenlock faunas are conspicuous for their 

 higher rather than lower range in North America, England, and south- 



