THE GLACIAL DRIFT. ^27 



ranges. Moreover, in this change of direction a constant position is 

 maimtained at right angles to the curvimg course of the Kettle Range. 

 The southeast bearing observed in Albion is an exception to this state- 

 ment, but this direction is evidently mei-ely a local one, since in the 

 country immediately east, as I am informed by Prof. Chamberlin, a 

 constant direction to the south or a little west of south is observed. 

 Other exceptions appear in two bearings observed in the country 

 about Baraboo, but these are from places in the valley between the 

 two quartzite ranges, and are evidently due, in some way, to their in- 

 fluence. In going northward through the eastern parts of Columbia 

 and Dane counties, though some increase in westing is seen, the gen- 

 eral directions are more nearly southwest. The outline of an area of 

 Trenton limestone that occurs in the adjoining corners of Dane, Co- 

 lumbia and Dodge counties has been given in Fig. 54. l^orth of 

 Columbia county the linear topography continues into Green Lake 

 county, but further west is not marked, nor are striee often to be ob- 

 served. 



In considering the origin and directions of travel of the erratics 

 and pebbles of the drift, we notice at once two classes of these mate- 

 rials, those that have been carried but short distances comparatively, 

 and whose exact place of origin may often be ascertained; and those 

 that have traveled all the way from the Lake Superior country, and 

 whose homes can generally be only roughly guessed at. It is from 

 the first class of bowlders that we can get our best ideas of the direc- 

 tions of the drift movement, not only because of the certainty of their 

 places of origin, but because they have probably moved in more direct 

 lines than those that have come from great distances. The following 

 are a few facts in regard to the first class of erratics! In the eastern 

 sections of the town of Deerfield, T. 7, K. 12 E., Dane county, are 

 many bowlders of a bluish-gray, flinty quartzite. associated with 

 others of a coarse quartzite conglomerate, both having evidently come 

 from the mounds of Archtean quartzite that rise through the St. 

 Peters sandstone on Sees. 34, 35 and 36 of the town of Portland, T. 

 9, E. 13 E., Dodge county. The distance traveled is from 9 to 14 

 miles, and the direction of travel S. 25°- 30° W., coinciding closely 

 with the directions of the topographical lines. On the top of the hill 

 just west of the depot at Lodi, Columbia county, S. E. qr. of the IST. "W. 

 qr. of Sec. 27, T. 10, E. 8 E., is a bowlder some eight feet high, of hard, 

 brownish sandstone, having a vitrified or qnartzitic weathered crust. 

 Four to six miles due east, on top of the high prairie of Arlington, 

 are five small patches of St. Peters sandstone, the rock of which has 

 characters exactly resembling those of the bowlder at Lodi. Midway 



