36 THE ROYAL SOCIETY OF CANADA 



tion of Dakota, a similar series of passage beds, shows that the Edmon- 

 ton formation was laid down before the Lance. To quote from 

 Barnum Brown, ^ "The vertebrate fauna is distinct from that of the 

 Lance and a few species are common to the two formations. Most 

 of the Edmonton genera are structurally more primitive than those of 

 the Lance and several genera not found in the Lance are common to 

 the Judith River. The faunal facies, as a whole, is intermediate, but 

 closer to that of the Judith River formation than to the Lance." 



In considering the shallowing of the sea during the deposition of 

 the Edmonton beds the question of a shore line in the south when the 

 Lance beds are now found may be discarded and the stages of the 

 retreat may be considered as consisting of at least two. The first, a 

 period in which the western margin was fringed with brackish water 

 deposits recognized in southwestern Alberta as the St. Mary forma- 

 tion increasing in thickness to the north and northeast in the Edmonton 

 formation. Probably these formations are bordered to the west by 

 continental deposits containing a Tertiary flora which would be the 

 earliest Eocene in the district. In southern Saskatchewan there is no 

 trace of these deposits so that the contemporary deposits are marine. 

 This first period terminates in a withdrawal of the sea to a line which 

 runs south of the Cypress Hills and then somewhat north of east to 

 probably south of Moosejaw. From this line southward, the deposits 

 of the second stage of the retreat are found in the Lance beds. The 

 limits of this second phase through the western States can not be 

 discussed in this paper, mainly from lack of information, but it is 

 quite possible that in the vicinity of the mountains both stages of the 

 retreat may be recognized. The absence of the Lance formation 

 beneath the Cypress Hills and along the northern flanks of the Wood 

 mountain plateau, and the presence of doubtful Lance beds at the 

 western and southern portion of the Wood mountain plateau, and a 

 possible Lance formation beneath the Fort Union of Estevan, are the 

 determining factors in the above statement that the first period of 

 retreat terminated by a steady withdrawal to a line from south of the 

 Cypress Hills to a point probably east of Moosejaw. 



The bed which is taken as a lower member of the Tertiary is the 

 white clay bed which is conspicuous in the Cypress Hills overlying 

 marine sands. In the Willowbunch area, east of Wood Mountain 

 Plateau, Longitude 105°, this white band is at the base of the sections 

 exposed, but the base of the formation seems there to be possibly 

 lower judging by the fact that the border is mapped by Bruce Rose as 

 at some distance both north and west. To the west along the 49th 



1 Bull. Geol. Soc. of America, Vol. 25, 1914, p. 374. 



