166 Mr. Hopkins on the Mechanism of Glacial Motion. 



place in a homogeneous mass; but if, on the contrary, the 

 planes ofless cohesion approximate to those directions in which 

 J'=.0 {i.e. to those perpendicular to the open fissures), they 

 will have no influence on the directions of the lines of fracture, 

 for such fractures must either be formed exactly along the hnes 

 ofless cohesion, or in the directions in which/'is the greatest. 



26. It may at first sight appear paradoxical that there 

 should be the same tendency to fracture the mass by one par- 

 ticle sliding past another in a transversal as in a longitudinal 

 direction ; it is, however, unquestionably true. The difference 

 between the two cases does not consist in any difference of 

 conditions previous to fracture (or, therefore, in the tendency 

 to produce fracture), but in the motion which may subsequently 

 take place in the respective directions. The relative motion 

 of particles on opposite sides of the transversal plane of frac- 

 ture, will be limited to that which is just necessary to relieve 

 the previous constraint of the mass ; while the relative motion 

 subsequent to a longitudinal fracture may be continued to any 

 extent by the continued tendency of the mass on one side of 

 such fracture to move faster than that on the opposite side. 



Whether a continuous portion of ice of any considerable 

 extent be ever fractured in the manner last considered, is un- 

 certain, for no accurate observations have yet been made on 

 the subject; but there can be no doubt that, in those portions 

 of a glacier where it is much dislocated, the sliding of one 

 portion past another must frequently take place. Such is 

 doubtless the case along the dislocated flanks of glaciers, and 

 thus the motion of the central part is much facilitated. If a 

 glacier, continuous throughout its whole extent, could be 

 formed, and then subjected to the existing conditions of gla- 

 ciers, I have no doubt that the first fractures would be ira?iS' 

 versal, produced by the tension (R) ; and that after the mass 

 had thus become considerably dislocated along its sides, the 

 sliding of one finite portion past another would also take place; 

 after which any internal constraint which might be superin- 

 duced by the peculiar motion of the glacier, would be relieved 

 both by transverse fracture and by the sliding of one part past 

 another. 



27. If a column of glacial ice, of which the unextended 

 length should = I, would just bear to be stretched to the 



length /' before it would break, the ratio —j— might be taken 



as a measure of the extensibility of glacial ice. If the exten- 

 sion should be produced within the comparatively short time 

 of making an ordinary experiment of the kind, this ratio would, 

 I think, unquestionably be very small; but if the extended 



