42 F. H. Lahee— Crescentic Fractures of Glacial Origin. 



itself a nearly flat stoss slope which has an inclination of about 

 10° northeastward (in a direction, N.65°E.). Upon all these 

 glaciated surfaces there is a high polish, and numerous, nearly 

 parallel, fine striae trend N.f °-10°E. It is a conspicuous fact 

 that the fractures occur on the northern sides of the roches 



Fig. 1. Section through quartzite showing glaciated surface (a-b), direc- 

 tion of ice movement (arrow), and seven crescentic fractures, passing into 

 the rock from a-b. 



moutonnees, and, further, that they are particularly well devel- 

 oped near the crest of the hill. 



The crescentic fractures themselves are grouped in sets, or 

 series, in which the separate members succeed one another at 

 short intervals* in the direction of ice motion, that is, parallel 

 to the striae. Each fracture is concave forward. f Its trac- 

 ing on the surface of the rock is an hyperbolic-curve^: of which 

 the transverse axis is also the axis of the series to which that 

 crack belongs (and therefore the general direction of ice 

 motion). The asymptotes of the curve usually form an angle of 

 about 90°. In any given set all the cracks are not of equal 

 length; there are many short ones for each long one. Hence 

 they are most numerous near the common axis (tig. 2). 



the position of the axis is not infrequently marked by a 

 stria, or even by a shallow groove which may be as much as 

 half an inch wide ; but such evidences of actual abrasion are 

 not always present. 



At the surface of the outcrop, and near the axis, the fracture 

 is distinct and clean-cut and it passes directly through those 

 grains of the rock that may happen to be in its course (iig. 1) ; 



* In one case 60 to the linear inch, parallel to the striae, were counted. 



f We use forward meaning with the ice motion. Andrews, Chamberlin, 

 and Gilbert, noted that crescentic cracks are concave forward and suggested 

 this as a criterion for determining glacier motion. They also observed that, 

 in this respect, these cracks resemble chatter-marks. 



In the present locality the shape of the roches moutonnees, the ' drag -lines ' 

 on the lee sides of hard rock obstructions, and other features, are sufficient 

 evidence for inferring that the ice advance was from north to south, and that 

 the late Pleistocene local glaciers, which Hitchcock (Geol. of N. H., Ill, pp. 

 233-238) describes as flowing north and west from the White Mountains, 

 either did not traverse Northey Hill, or, if so, were no more effective in caus- 

 ing striation and the usual accompanying phenomena here than they were in 

 the region over which he shows they flowed (Bethlehem, etc.). 



\ And therefore not truly crescentic. We use the term crescentic, how- 

 ever, following the rule of precedence. 



