508 E. O. ULKICH REVISION OF THE PALEOZOIC SYSTEMS 



(2) "Matching" of species and genera. — This is the first step in corre- 

 lation by fossils. The degree of similarity exhibited by geographically 

 separated faunas is usually proportionate to their respective ages. If 

 great, then the evidence is provisionally accepted as indicative of essential 

 contemporaneity. But the result of mere matching of faunas, however 

 close, is, of itself, never conclusive. For instance, there is a fauna of 

 about 80 species in the basal part of the Liberty Hall limestone in west 

 central Virginia. The fauna includes species of several genera usually 

 regarded in America as Cambrian. A much larger proportion of the 

 fauna consists of species closely allied to or identical with Chazy types. 

 Others are unknown elsewhere. Finally, there are some species whose 

 evidence is exact and fully in accord with the stratigraphic evidence on 

 which the bed is determined to be younger than Chazy and older than 

 Trenton. Obviously, simple matching in this case might lead to serious 

 error. It shows, also, that careful discrimination of the faunal evidence 

 and elimination of the non-essential factors are sometimes required and 

 always desirable before reaching final conclusions. This is especially 

 necessary in cases of reappearing faunas like the Spergen. The species 

 of this fauna have a definite zonal value only in the Spergen zone itself. 

 Hence, in correlating such later horizons as the Sainte Genevieve and the 

 Tribune, the Spergen constituents of their respective faunas must be 

 eliminated from the list of characteristic species and reliance confined to 

 the remaining more exactly diagnostic forms. 



(3) Dominant species. — Individual dominance of species is not a re- 

 liable test of the chronological significance of local faunal aggregates, 

 because (a) the list of dominant species necessarily includes the relatively 

 vigorous types which, under average conditions, are likely to be more 

 adaptable, longer lived, and correspondingly less definite in their time 

 relations than their weaker and more sensitive associates; and (b) the 

 numerical representation of any and all organisms varies from place to 

 place according to local changes in environment, and probably on account 

 of other more fortuitous circumstances, so that a given species may be 

 found at one locality in sufficient abundance to place it at the head of 

 the list of dominant fossils, while its rarity in beds of exactly the same 

 age at another place may exclude it entirely. 



The first case is illustrated by such vigorous, long-ranging, though 

 doubtless too loosely conceived species as Leptcena rhomb oidalis, Rafines- 

 quina alternata, Plectambonites sericeus, Dalmanella testudinaria, Platy- 

 strophia biforata, Isotelus gigas and Calymene senaria. All of these 

 are found at intervals through long geological terms, and when they 



