THE SUCCK8.SI0X OF DEPO.^ITS. 



69 



preceded the re-elevatioii of the Saxicava sand * and it is 

 even possible tliat some may have lieen jJaeed in tlieir 

 present positions in tlie ])()St-ghK'ial subsidenee, of which 

 there is evidence on l)oth sides of tlie Athmtic. S(jme 

 belong to lake margins (»f jtost-glacial date. Thns no 

 general statement can safely be made respecting these 

 erratics, but each group or belt must l)e studied with 

 reference to its local associations and the source of the 

 nijiterial, as well as with reference to the pro1)able stage 

 in the various continental subsidences and elevations to 



Fig. 3.— Teiraecs at L'Aiisu a LcMip, near Taaousac. Lower terrace, clay ; upjier, sand. 



which it belongs. The aminq^tiim that all loiddcr-drift 

 may belong to one period is a fertile souree of error, and 

 though many important observations on the subject have 

 been made by Spencer, Dr. G. M. Dawson, Chalmers and 

 others, there is an almost unlimited field for detailed work- 

 in this direction. 



A still farther complication arises here from the pro- 

 bability of differential elevation, whereby the relative 

 levels have been changed in different parts of the l»leisto- 



* McGee refers to this in connection with the "Columbia" deposits 

 of the Appalachians. 



