298 H. S. WILLIAMS PERSISTENCE OF FLUCTUATING VARIATIONS 



evolutional modification^ it seemed necessary to trace the genus through 

 more than a single stage of its life history, for the reason that any 

 morphologic differences expressed by a genus under a single set of 

 environmental conditions can be interpreted only as fluctuating varia- 

 tions within the limits of specific characters and may have nothing to do 

 with evolutional modifications. 



Effect of geographic Distribution 



In order to be sure that geographic distribution did not enter in to 

 complicate the evidence, it was necessary to collect the samples from a 

 series of successive beds within a limited area within the confines of the 

 same geologic basin. 



Artificial selection of Evidence 



Further to insure that no artificial selection of evidence take place, 

 uniform conditions of collecting the evidence and a very thorough search, 

 and the subjecting of every trace of the evidence to examination were 

 necessary conditions. 



BiOMETRic Method 



And, finally, because of the elastic and indefinite nature of descriptive 

 terms applied to form, it was necessary to reduce the characters under 

 examination to mathematical terms, so that they could be accurately 

 stated and compared. I made all measurements myself, and with the 

 same instruments, in order that whatever personal equation might enter 

 into the statistics the errors would be reduced to a minimum by being 

 uniform for all the statistics. 



Location AND geologic Horizon of Specimens 



It has taken a number of years to satisfy all these conditions (I started 

 the investigation in 1881), and in August last I began, for the first time, 

 to make the measurements, the results of which I now have to report. 



The geological sections from which the specimens were obtained are all 

 within an area in central New York within a radius of 50 miles from 

 Ithaca as a center. 



Stratigraphically, the zones range from the middle of the Hamilton 

 formation, through the G-enesee, Tully, Portage, and to the top of the 

 Cayuta member of the Chemung formations, presenting a total thickness 

 of approximately 2,300 feet, which is at least 1/100 part of the thickness 

 of the whole sedimentary beds of the earth's crust from which distinct 



