620 INDEX TO THE STRATIGRAPHY OF NORTH AMERICA. 



be borne in mind is that structurally and faunally, so far as known, the Knoxville is a unit. 

 It is true that there is a gradual change in the fauna from the lower to the upper beds, but 

 there is no distinct break that would jiastify the reference of one portion to the Jurassic and 

 another to the Cretaceous. There are some elements of the fauna, such as Belemnites tehamaensis 

 Hoplites storrsi, and some of the Turbinidas, that resemble Upper Jurassic types. AuceUor 

 piochii belongs to the same general type as A. mosquensis, which has usually been referred to 

 the Jurassic, and it is somewhat closely related to the AuceUas " of the Mariposa beds, but Prof. 

 Pavlow informs me that Aucellse very similar to A. mosquensis occur in the Neocomian of Russia 

 also. These resemblances, which will be more fully mentioned in the description of the species, 

 are not considered of sufficient importance to counterbalance the evidence of the Cretaceous age 

 of the entire series. Some Jurassic elements are naturally to be expected in the Lower Neo- 

 comian, and it is a well-known fact that in Europe several species of ammonites pass up from 

 the Jurassic to the lowest Cretaceous beds. 



Prof. Hyatt's opinion as to the Cretaceous affinities of the Knoxville ammonites has been 

 cited on a previous page. That the Upper Knoxville beds containing AuceUa crassicoMs are 

 Neocomian is shown by the resemblance of its fauna to that of the Petschora beds, now gener- 

 ally admitted to be Neocomian, and by its stratigraphic and faunal relations with the immedi- 

 ately succeeding Horsetown beds, which have been compared with the Gault by everyone who 

 has studied their fauna.^ The few fossil plants that have been obtained from the Knoxville 

 and the Horsetown also tend to shew that these two formations are closely connected, and these 

 plants are mostly either identical or closely related with species in the Lower Cretaceous Potomac 

 formation of the eastern United States and in the Trinity beds at the base of the Comanche 

 series in Texas." 



From aU these Hnes of evidence the conclusion is reached that the entire Knoxville series 

 is of Neocomian age. 



In 1910, as this manuscript is being finished, Stanton states orally that the 

 preceding quotation expresses the views he stills holds. 



J. Perrin Smith ''^^ does not altogether agree with Stanton, as appears from the 

 following statement : 



The Cordilleran revolution began in the Great Basin sea in the middle of the Jurassic, when 

 that body of water, after many vicissitudes, finally went dry, and has never since been covered 

 by salt water, although in later ages Tertiary and Quaternary lakes have been scattered over 

 its dead basin. 



This elevation culminated, in late Jurassic time, in the upturning and metamorphism of 

 the Triassic and Jurassic sediments of the Sierra Nevada and the Franciscan beds of the Coast 

 Ranges. * * * 



It is probable also that the Cordilleran revolution was something more than a mere ero- 

 genic disturbance, for it marks a change from the warmth of the Middle Jurassic, with its cycads 

 and reef-building corals, to the cooler epoch of the Upper Jurassic, with its scanty boreal fauna. 

 The Middle Jurassic was oi tropical type, from Mexico to Alaska, and uniform up to Franz 

 Joseph Land. The Upper Jurassic, on the other hand, was of boreal type from the Arctic 

 region down as far as California, and for a short epoch in the Portland these conditions extended 

 down as far as Mexico. 



After this mountain-making epoch near the close of the Jurassic, the sea again encroached, 

 on the uplifted area, and the Knoxville sediments were laid down on the western border of the 



o See comparieonB bV Prof. Hyatt, Bull. Geol. Soc. America, vol. 5, pp. 404^07. 



6 It is not probable that the limits of the divisions recognized in the California Cretaceous coincide exactly with 

 those of the European subdivisions of the Cretaceous. It has been shown (Bull. Geol. Soc. America, vol. 5, pp. 445- 

 449) that the Horsetown beds probably include the lowest Cenomanian, and at their base they may contain strata, 

 older than the Gault. 



" These piants are the only means by which the Lower Cretaceous of Texas and California can be compared. Sa- 

 far as known, their invertebrate faunas have nothing in common. 



