OSCILLATORY CHARACTER OF CONTINENTAL SEAS 349 



sean, and between the Chemung at the top of the Devonian and the next 

 overlying Waverlyan ("Mississippic") system which I begin with the 

 Chattanooga shale in Alabama, middle Tennessee, and Arkansas, and 

 with the Cleveland shale in Ohio. In all of these instances, indeed be- 

 tween all the systems and many of the divisions of less rank, the evidence 

 of complete withdrawal of seas from at least the eastern part of the conti- 

 nent seems indisputable. 



Time values of systems. — In estimating relative time values of sys- 

 tems it is essential to consider the probable duration of the non-accessible 

 deposition intervals as well as the parts represented by marine sediments 

 in continental seas now exposed to view. This introduces an element of 

 great uncertainty into estimates purporting to give the probable duration 

 in years of periods and eras, and positively negatives all attempts that 

 fail to take these inten^als into account. That the principle is generally 

 recognized in practice so far as continental marine deposits are concerned 

 is shown by tlie fact that the time values of formations and groups of 

 formations are determined by the fullest available sedimentary record 

 and not by records which by comparison we know to be much less com- 

 plete, or indeed but fragmentary. Including these intervals of non-acces- 

 sible deposition the total of geologic time is greatly augmented. The 

 time value also is increased proportionally for those periods which are less 

 fully represented by deposits in continental seas than are other periods 

 during which the average elevation of the continent above sealevel was 

 less and in which, consequently, the available depositional record is less 

 defective. Without taking this factor into consideration, and believing 

 that the respective time values of the successive periods are not greatly 

 unequal, the known deposits of the proposed Waverlyan and Tennessean 

 systems or periods, for instance, would scarcely be sufficient to Justify 

 their separation. Large parts of these periods, namely, are thought to be 

 included in long intervals of inaccessible record that preceded in tlie one 

 case the Warsaw, in the other the Pottsville submergence. Both of these 

 cases are preliminary^ stages of the emergent phase of the N'orth American 

 continent that culminated in the Permo-Triassic interval. Following the 

 Triassic the balance of continental oscillation again favored submergence, 

 this particular cycle attaining its maximum marine invasion in the Cre- 

 taceous. 



Except to say that I believe the age of the earth, beginning with the 

 Cambrian, is much greater than 70,000,000 years, the amount that since 

 Walcott's estimate in 1894^® is being rather generally accepted as satis- 



Proc. Am. Assoc. Adv. Sci.. vol. 42, 1894, pp. 129-169. 



