440 C. SCHUCHERT PALEOGEOGRAPHY OF NORTH AMERICA 



in the Lower Kichmond formation never more than one foot thick. C. 

 erratica is always diagnostic of the eastern Lorraine. Triplecia ortoni is 

 restricted to a limestone zone, never more than 25 feet in thickness, ex- 

 tending from Oklahoma to Ohio, while a closely related species occurs in 

 a similar zone on the island of Anticosti in the Saint Lawrence gulf. The 

 higher Clinton, from Anticosti to Alabama, may be determined by a single 

 brachiopod, Anoplotheca liemisplierica, which also identifies a very similar 

 horizon in northwestern Europe. Rhynchotrema capax in two varieties 

 defines the Kichmond formation from El Paso, Texas, to Manitoba, and 

 from the Big Horn mountains to Ohio, and even to Anticosti. Spirifer 

 hung erf or di denotes the L T pper Devonic throughout western America, 

 from the Arctic to Bisbee, Arizona, and in Asia and Eussia as well. These 

 are not exceptional cases, and many more can be cited. When fossils are 

 carefully collected from bed to bed, it is nearly always found that certain 

 combinations of species called faunides may be depended upon to indicate 

 unvarying zonal or time values within a subprovince. Sometimes the 

 latter is of very wide extent, as in the case of the Upper Kichmond forma- 

 tion just mentioned, while in other cases they are greatly restricted. As 

 a rule, the more common species can not be relied upon for the determina- 

 tion of limited horizons, especially the ubiquitous plastic forms among 

 brachiopods. For instance, Atrypa reticularis, Leptcena rhomb oidalis, 

 Dalmanella testudinaria, Plectambonitcs sericeus, Rafinesquina dlternata, 

 Pentamerus oblongus, etcetera, have little value for interprovincial corre- 

 lation, and none at all when loosely identified, as is the almost universal 

 practice among paleontologists. 



Correlation between provinces is far more difficult, but when large col- 

 lections are at hand the general faunal facies, together with a few iden- 

 tical or closely related species, will usually enable a paleontologist to fix 

 upon a fairly definite time. Exact correlation that may be proved be- 

 comes impossible only in newly discovered areas, as is the case at present 

 for Alaska. Even there, however, the faunas are now taking on deter- 

 minable provincial relationship with those farther south. 



Barriers. — In the Ignited States there- are often two or more provincial 

 faunas of the same geologic age, apparently having geographically contin- 

 uous strata that are wholly different, there being but few species common 

 to them. As an example may be cited the Middle Devonic coral faunas in 

 Kentucky, Indiana, Ohio, Xew York, Ontario, and Hudson bay. Con- 

 trasting this well known assemblage with the less abundant coral repre- 

 sentation in Iowa and Missouri and in the entire country west to the 

 Pacific, it is seen that while there is a time expression common to both 

 provinces, there are practically no identical species. The eastern assem- 



