I 



OSCILLATORY CHARACTER OF CONTINENTAL SEAS 345 



Intersystemic intervals. — It will be observed that the withdrawal of 

 the seas is represented as complete at the close of each of the successive 

 periods. By complete withdrawal is meant a condition like the present — 

 that is, a retreat of the waters to the marginal shelf on which the record 

 is buried beyond present possibilities of investigation. The supposed 

 duration of these intervals of complete withdrawal is indicated by the 

 relative width of the unshaded spaces at the margins of the diagrams. 

 The break between the systems is great where the space is wide on all of 

 the four invading sides, and where it is narrow on any side the interval 

 of non-accessible record is thought to be relatively short. It seems to 

 have been shortest between the Silurian and Devonian and not much 

 longer between the Ordovician and Silurian and the Devonian and Waver- 

 lyan. It is thought to have been longer between all the others. As or- 

 ganic evidence favoring this hypothesis I would point out the fact of 

 close relationship exhibited by the faunas of, respectively, the late Ordovi- 

 cian and early (Richmond) Silurian, the late Silurian and early De- 

 vonian, and late Devonion and early Waverlyan. These were the periods 

 of greatest submergences. They are also the systems that, despite their 

 great wealth of fossils, have given the greatest trouble to bound, so long 

 as faunal evidence alone was considered. The relatively gradual passage 

 of the older to the newer of these faunas should be contrasted with the 

 sharper life breaks separating all the other Paleozoic systems. 



When the intersystemic interval is short a large part of its record has 

 sometimes been made in greatly reduced continental sea basins and in 

 land deposits in areas adjacent to relatively permanent "positive" re- 

 gions. Examples would be the late Cincinnatian marine deposits in the 

 southwestern part of the Ohioan province and contemporary land de- 

 posits ( Oswego and Juniata sandstones ) in Pennsylvania and New York ; 

 marine Cayugan sediments in the middle Appalachian Valley region and 

 mostly lake or land deposits farther inland; late Devonian of marine 

 origin in various parts of the country and land deposits of corresponding 

 age in New York and Pennsylvania. When the interval is long the 

 marine record is wholly confined to inaccessible marginal areas and the 

 ]ocal land deposits that may have been made in the earlier part of the 

 interval were in part or whole removed again in the course of base- 

 leveling before the succeeding submergence of the next period could cover 

 and preserve them. The Pleistocene land record, for instance, is rela- 

 tively new. Under a drier climate it might all be removed before another 

 marine submergence solidifies and preserves it. At best only remnants 

 could be preserved. 



XXIV— Bull. Geol. Soc. Am., Vol. 22, 1910 



