312 G. K. GILBERT — CRESCENTIC GOUGES ON GLACIATED SURFACES 



the glacier is moving in the direction indicated by the arrow AA; that 

 the rock floor is indicated by HH; that the boulder FF is either in con- 

 tact with or near the rock floor. Under these conditions the direction of 

 the greatest stress would be indicated by the arrow BB, being the result- 

 ant of the weight of the glacier and the pressure behind the moving mass. 

 Under these conditions there are powerful shearing stresses in the direc- 

 tions BC and BD. These stresses are greater adjacent to the boulder 

 because it is a rigid body and is able to transmit forward close to the 

 rock floor the pressure of the ice about it. At the place where there is the 

 most rapid change in. the amount of compression rupture takes place. 

 Whether the rupture occurs in both the horizontal and vertical directions 

 will depend upon circumstances, which will largely depend upon the 

 shape of the rock surface and the position of the boulder and its shape. 

 When the vertical rupture takes place alone you have the crescentic 

 cracks, when horizontal rupture takes place, followed instantaneously by 

 the vertical rupture, you have the crescentic gouges." 



Bhythm 



A full development of the hypothesis would* include also a discussion •>£ 

 the occurrence of the gouges in series, and this likewise requires the ex- 

 pert knowledge of the elastician. If I again venture a suggestion it is 

 largely in the hope of exciting his interest. There can be little question 

 that each series of gouges represents a mechanical rhythm of some sort. 

 In a large group of mechanical rhythms, including many in which fric- 

 tion plays a part, a force uniformly applied accumulates strain and stress, 

 which are relieved in some catastrophic manner whenever they reach a 

 certain limit. In the present case the conoid rupture is a catastrophic 

 event relieving some of the internal stresses of the bed-rock. The jar, or 

 miniature earthquake, occasioned by it and radiating from the point of 

 rupture may be supposed to overcome frictional resistance between glacier 

 and rock and cause a sudden slipping along their contact surface, thus 

 relieving the frictional strains and stresses for some distance in all direc- 

 tions. The boulder instantaneously moves forward to a new position with 

 reference to the rock bed, and the gradual renewal of deformation and in- 

 ternal strains is begun. This line of inference leads to the difficult ques- 

 tion whether the sudden forward movement covers only the fraction of an 

 inch, or whether it may be of the order of magnitude of the interval be- 

 tween gouges — from a few inches to several feet. If it is very small, 

 then the determination of the gouge interval remains as one of the ob- 

 scure factors of the hypothesis. In a general way the gouge interval is 



