MOOK, 8TLWY OF THE MORRISON FORMATION 163 



character of the aquatic molliisca to preclude the possibility of their hav- 

 ing lived and developed in smaller lakes. After a personal examination 

 of the localities at Green Eiver, Utah, at Grand River in western Colo- 

 rado, Canon City and Morrison in eastern Colorado, Como and Sheep 

 Creek in southern Wyoming, at the Spanish Mines in eastern Wyoming, 

 along the Bighorn Mountains in central Wyoming, about the Black Hills 

 in South Dakota and in the country near Billings in southern Montana, 

 in all of which localities the Atlantosaurus heds are exposed and exhibit 

 in more or less abundance, the remains of those dinosaurs which are 

 characteristic of them, I am convinced that neither the character of the 

 vertebrate fauna nor the facts of stratigraphy at any one of these places 

 can be taken as affording anything like conclusive evidence of the presence 

 of a great body of water. At several of these localities, however, the oc- 

 curence at intervals of sandstones showing frequent examples of cross- 

 bedding, ripple-marks and even occasionally exhibiting footprints is con- 

 clusive proof that such sandstones had not their origin in the midst of a 

 great lake, while the j)resence almost everywhere of the remains of terres- 

 trial reptiles and less frequently of mammals tells only too plainly of an 

 adjacent land-mass. In all this region I know of no locality where any 

 considerable extent of the Atlantosaurus beds occurs, in which remains of 

 quadrupedal, terrestrial dinosaurs have not been found. . . . An 

 hypothesis, which it appears to me is far more reasonable and more nearly 

 in accordance Avith the facts as we now know them, is to consider this 

 region as presenting in late Jurassic and early Cretaceous times the ap- 

 pearance of a low and comparatively level plain, with numerous lakes, 

 both large and small, connected by an interlacing system of river chan- 

 nels." 



Chamberlin and Salisbury (1907, 7) assign a fluviatile origin to tbe 

 Morrison formation. 



Lee (1915, 2) considers the Morrison to be largely fluviatile in origin. 



Discussion" of Previous Theories of the Origin of the Formation 



The theory of deposition of the Morrison in a great lake, as advanced 

 by C. A. White, does not seem to be supported by evidence now available. 

 The following list of characteristics of beds of lacustrine origin has been 

 given by Johnson (1903, 8) : 



1). No great variations in texture and composition in vertical section. 



2). IsTo beds of conglomerate. 



3). ISTo marked and sudden variations in respect to the thickness and 

 areal extent of the component beds. 



