G. H. Stone — Drift Scratches in Maine. 147 



accurate system of measurements than has heretofore prevailed. 

 A full discussion of this subject is reserved for another occasion. 

 For our present purpose it is assumed that in order to prove 

 by means of glacial scratches, that the ice at a given place 

 moved in different directions at different times, it is needful to 

 prove that there are intersecting series of scratches at that point. 



The following instances of local changes in the direction of 

 the ice-flow in Maine have been reported. In the Maine Geo- 

 logical Eeport for 1861 (p. 262), Professor C. H. Hitchcock 

 reports intersecting drift striae at Chamberlain Lake in northern 

 Maine, and perhaps his observations on the drift of the upper 

 Saint John Yalley may indicate local glaciation. Professor 

 Hitchcock has also described a roche mouionnee in York County, 

 Me., on the stoss side of which the drift scratches bear south- 

 eastward, while on the lee side they bear south westward. 

 Professor G. L. Vose was the first (American Naturalist, vol. 

 ii, p. 28) to report the signs of a local glacier in the Andros- 

 coggin Valley. This glacier extended from the White Moun- 

 tains eastward to near West Bethel, Me., as is proved by the 

 moraines and by the glacial scratches which, near the bottom 

 of the valley, are nearly parallel with the course of the valley, 

 while the glaciation of the higher hills and the surrounding 

 region is nearly at right angles to this direction. These ob- 

 servations regarding the local Androscoggin glacier, the writer 

 has carefully verified. The other localities named have not 

 been visited by me. Professor F. C. Robinson of Bowdoin 

 College, has also informed me that he has found intersecting 

 drift scratches on the bank of the Penobscot River, a few miles 

 west of Chesuncook Lake. 



During the past summer the writer has discovered three 

 other localities showing intersecting drift strise. That the 

 secondary glaciation was not sooner discovered is due to the 

 fact that the slates east of Waterville weather quite readily. 

 The later scratches are shallow, and long after they have been 

 obscured by weathering, the earlier and deeper ones remain 

 quite distinct. This gives an appearance of freshness to rock 

 exposures, which is very deceptive. 



1. The SebasticooJc Valley Locality. 



The Sebasticook River flows from the east into the Kennebec 

 River at Waterville. Numerous observations of the drift striae 

 on both sides of the Sebasticook show that there is an area 

 several miles long in the towns of Winslow, Benton and Clin- 

 ton, where there are two and in a few places three superposed 

 systems of scratches. In a few places very near the Sebasticook, 

 the later scratches are very confused, pointing in many direc- 

 tions, whether from berg drift or ice gorges of the river is yet 



