THE FACTS 



387 



erh Europe, and for their conspicuous appearance in the Silurian faunas 

 of England and Gotland, Bohemia and Podolia, but their absence from 

 the corresponding Silurian faunas of interior America, though having 

 representation in several cases in faunas of recognized higher horizons. 



Thus far the conclusions reached seem to rest on what we may regard 

 as established facts, the certitude of which can be verified by examination 

 of the specimens. 



When we go further and attempt to interpret the presence or absence 

 of certain so-called identical fossils into time-relations of the formations 

 containing them, and still further try to reconstruct from them the geog- 

 raphy of the earth's surface at some particular stage of geological' time, 

 we are obliged to bring in assumptions which do not find their verifica- 

 tion in observed characters of the fossils, but are based on biological 

 hypotheses and even metaphysical conceptions we have been accustomed 

 either to neglect or take for granted rather than examine with the same 

 care we devote to identification of species. 



It is to these hypotheses and assumptions that I particularly wish to 

 call attention in what follows : 



Part II. Problems involved in correlating Fossil Faunas 



GENERAL CONSIDERATIONS 



What has previously been said concerns the observed facts in the case^ 

 In the present discussion I will ask you to take it for granted that they 

 are correctly stated, because for the purposes of this paper a hypothetical 

 case would serve as well as a real one, provided the conditions assumed 

 were all natural and in accordance with observed facts. 



The chief purpose I have in writing this paper is to call attention to 

 the rather remarkable number of inferences we are accustomed to draw 

 in framing our ideas in paleontology, stratigraphy, and paleogeography, 

 which, though they appear to rest on what we regard as established, ob- 

 served facts, may have very little or no logical connection with them, or 

 may be so loosely or so irrationally applied as to render the conclusions 

 unsound or positively false. 



The case is like this : I have examined the fossils taken from the Ed- 

 munds formation of the Eastport quadrangle, Maine, and listed them 

 under scientific names. Whenever I could find a name which had been 

 already applied to the characters observed in species of the Edmunds 

 fauna I used it. If new characters were discovered T have described the 

 form under a new name. In making comparison with the fossils of for- 

 mations already described, I have used what may be called the statistical 



