332 Mr. Hopkins on the Mechanism of Glacial Motion. IV. 



appeared on each side of the central portion, but in a position 

 entirely different from those above-described. When first 

 visible their common direction made an angle of 15^ or 16° 

 with the axis of the trough, whereas the systems of lines when 

 the mass was more compact, were at an angle of 45° with the 

 axis. To ascertain whether there was any tendency in the 

 particles on opposite sides of one of these new lines of disloca- 

 tion to slide past each other, I proceeded as in the former 

 case, drawing a short line on the surface perpendicular to one 

 of the dislocations, and continuing the central motion. The 

 short line drawn on the surface immediately became a broken 

 line at the point of intersection, proving the existence, in this 

 case, of the sliding motion, of which it will be recollected there 

 was no indication whatever in the experiment previously de- 

 scribed, when the mass was more compact. 



The lines now spoken of were not visible till the curvature 

 of the transverse loop had become much greater than in the 

 preceding case. On calculating the directions in which the 

 tangential action was a maximum, by the construction pre- 

 viously given (Second Letter, art. 17)} it appeared that those 

 directions made an angle with the axis about half as great as 

 that made by the lines of dislocation, when those lines first 

 showed themselves. Now in supposing, as I have in my pre- 

 vious letters, that the lines of greatest tangential action are 

 those along which separation planes would be formed, if formed 

 at all, I have tacitly assumed the resistance afforded by the 

 mass to their formation to be the same in all directions. It 

 would seem highly probable, however, that, cceteris paribus, 

 this resistance would be greatest in directions perpendicular 

 to the lines of greatest pressure, in which case dislocations 

 would take place in directions deviating 7^55 than the directions 

 of greatest tangential action from the lines of maximum press- 

 ure. Such was the case in the experiments just described. 

 The difference of position in the systems of dislocations, the 

 mass being compact in one case and soft in the other, was 

 owing, I conceive, to the mass yielding most easily in the 

 former case to the direct twrmal tension, and in the latter case 

 to the tangential action. It seems probable that such would 

 be the case ; and, in fact, in those experiments in which the 

 mass was softest, I doubt whether it was ever brought into a 

 state o^ tension at all, on account of its semifluidity, in which 

 case there would be no tendency at all to form open fissures, 

 as in the first set of experiments, while the formation of sepa- 

 ration planes would be facilitated. 



In the annexed diagram the stronger lines represent the 

 lines of dislocation in my experiment, when the mass was soft, 



