THE ERRATIC PHENOMENA. 407 



the other, and where they at times ascend steep hills, which they 

 cross at right angles along their northern slope, even when the 

 southern slope, not steeper in itself, faces the south with rough 

 escarpments. 



The scratches and furrows, though generally running north and 

 south, and deviating slightly to the east and west, present in various 

 places remarkable anomalies, even in their general course along the 

 eastern shore of the lake. Between Michipicotin and Sault St. 

 Marie we more frequently see a deflection to the west than a due 

 north and south course, which is rather normal along the northern 

 shore proper, between Michipicotin and other islands, and from the 

 Pic to Fort William ; the deep depression of the lake being no doubt 

 the cause of such a deviation, as large masses of ice could accumu- 

 late in this extensive hollow cavity before spreading again more uni- 

 formly beyond its limits. To the oscillations of the whole mass in its 

 southerly movement, according to the inequalities of the surfaces, 

 we must ascribe the crossing of the straight lines at acute angles, as we 

 observe also at the present day under the glaciers, as they swell and 

 subside, and hence meet with higher and lower obstacles in their 

 irregular course between the Alpine valleys. 



In deep, narrow chasms, however, we find now and then greater 

 deviations from the normal direction of the strige, where considerable 

 masses of ice could accumulate, and move between steep walls under a 

 lateral pressure of the masses moving onwards from the north. Such a 

 chasm is seen between Spar Island and the main land opposite Prince's 

 Location, south of Fort William, where the furrows and scratches run 

 nearly east and west. But here also, there is no tumultuous disturb- 

 ance in the continuation of the phenomena, such as would occur if ice- 

 bergs were floated and stranded against the southern barrier. The 

 same continuity of even, polished surfaces, with their scratches and 

 furrows, prevails here as elsewhere. The angles which these scratches 

 form with each other are very acute, generally not exceeding 10 ; 

 but at times they diverge more, forming angles of 15, 20 and 25. 

 In a few instances, I have even found localities where they crossed 

 each other at angles of no less than 30 ; but these are rare excep- 

 tions. It may sometimes be noticed that the lines running in one 

 direction form a system by themselves, varying very little from strict 



