502 
Buell—Bowlder Trains . 
over the whole area westward to the Crawfish river in Portland 
township. But in the ravines and on the steeper lower slopes 
of the same ridges in sections 5 and 6, Shields township, the case 
is different. Whenever in crossing these sections these lower 
levels are passed the deeply eroded quartzite material charac¬ 
teristic of this bowlder distribution appears. On the west line 
of section six about 200 of these were counted in a road¬ 
side fence whose material was gathered from a surface of perhaps 
two acres. In each of the half dozen ravines that lie to the 
east and west of this, similar amounts were seen. 
Beyond the township line on this north border the surface is 
less undulating and the quartzites are of rare occurrence, though 
in each road side gully from the south line of sec. 6, Shields, to the 
corresponding line of sec. 7, in Portland township, one or 
more local fragments appear, and an occasional bowlder is found 
on the surface. It is impossible to trace the south margin of 
this train across the area because the succeeding southward 
drift movements which crossed it and concealed its material 
over so large a portion of its surface commingled much of the 
underlying material with their detritus and redistributed it 
over their own broad fans. It is apparent though from the 
gradual declination of the north margin as traced across the 
area that the direction of ice movement in this part of its area 
was a little south of west, and that the center of the train 
crossed the north half of Portland township to the center of its 
west line. Quartzites appear in considerable numbers on the 
borders of ravines and on the steeper ridge slopes from sections 
12 and 13 to 18 and 19. fcouth of this line the more abundant 
and completely disseminated distribution appears to belong to 
later south westward trains from the same sources. 
The westward train from the Portland ledges is marked by 
less abundant surface distribution over the areas bordering the 
south line of Portland and York townships. Many blocks, how¬ 
ever, were found on the surface on the east line of sections 30 
and 31, Portland. Along the roadside on the south line of sec¬ 
tion 31, Portland, and 36, York, the same peculiarity of distri¬ 
bution was observed as upon the area west of the Mud Lake out¬ 
crops, namely, the absence of local drift on the crest of the 
