OSCILLATORY CHARACTER OF CONTINENTAL SEAS 371 



of distribution might not be so obviously fatal to Mr. Willis's suggestion. 

 But with this knowledge in hand, supplemented by the further fact that 

 the two boreal faunas alternate with a probably Pacific and a Gulf of 

 Mexico or an Atlantic fauna, all grounds for the proposition are obliter- 

 ated. According to the evidence, we are forced to the conclusion that 

 currents had nothing to do with the case. On the contraxy, we must 

 assume that these middle Mohawkian faunas represent four distinct 

 and in part, if not for the whole of each, successive invasions of shallow 

 basins on a nearly base-leveled, gently oscillating continental surface. 

 Further, of we admit partial contemporaneity of the boreal and southern 

 invasions, the final confluence of the basins must, eyen at its maximum, 

 as is shown by the continued integrity of the respective faunas, have been 

 very inconsiderable. Otherwise, a decided mingling of the faunas must 

 have resulted. Obviously, then, without free communication between the 

 basins, "channels across the continent, through which oceanic currents" 

 might have circulated, were impossible. 



Evidence of the graptolites on currents in continental basins. — -The 

 geographic distribution of the Eopaleozoic graptolites offers another 

 strong argument against the hypothesis of transcontinental currents in 

 the interior basins. I shall cite only the case of the Normanskill grapto- 

 lites because these flourished in the same period though earlier than the 

 "late Black Eiver and early Trenton" submergences here chiefly con- 

 sidered. Besides, the other graptolite faunas, especially the Levis, lead 

 to similar inferences. The Normanskill fauna, which comprises little 

 besides graptolites, occurs in great development in the shales of the 

 Levis and Athens troughs, respectively the northern and southern exten- 

 sions of the eastern parts of the Appalachian Valley. It is abundant 

 also in one of the shales of the similar Ouachita trough in Arkansas and 

 Oklahoma. The same fauna is found also in Great Britain and Sweden 

 and is recognized in Australia, on the opposite side of the globe. Obvi- 

 ously, this is a truly cosmopolitan pelagic fauna whose great dispersal is 

 owing chiefly or solely to transportation by oceanic currents. 



In North America the Normanskill graptolite fauna is confined to 

 strips of shale just west of the Chilhowee-Green' Mountain barrier of 

 Ulrich and Schuchert. The age of the shale has been variously estimated. 

 For many years it was thought to be the same as the Utica but now it is 

 known to be older. Euedemann and others regard it as middle Trenton 

 and perhaps most geologists, Willis among them, now view the Norman- 

 skill shale as a nearer shore phase of supposedly contemporaneous lime- 

 stone formations farther west. That it can not be this is proved con- 



