[539 ] 

 LXXIX. Proceedings of Learned Societies. 



ROYAL SOCIETY. 



[Continued from p. 4//.] 



June 18, 1863. — Major-General Sabine, President, in the Chair. 

 '^PHE following communications were read : — 

 -*- " Some Remarks appended to a Report on Mr. Hopkins's Paper 

 * On the Theory of the Motion of Glaciers. 3 " By Sir John F. W. 

 Herschel, Bart, F.R.S. 



A few remarks arising out of the perusal of this paper may per- 

 haps not be considered as out of place on the present occasion. They 

 are not meant as in any way impugning the author's views of the 

 laws determining the fracture and disruption of glacier masses, or 

 their application to glacier-phenomena in general, but in relation to 

 the somewhat mysterious process of regelation itself, and to those 

 generally recognized and most remarkable facts of the gradual con- 

 version of snow into more or less transparent ice, and the reunion of 

 blocks and fissured or broken fragments, under the joint influence of 

 renewed pressure and of that process (whatever its nature), into conti- 

 nuous masses. If regelation be really a process of crystallization, it 

 seems exceedingly difficult to imagine how the molecules forming the 

 cementing layer between two juxtaposed surfaces can at once arrange 

 themselves conformably to the accidentally differing axial arrange- 

 ments of those of the two surfaces cemented. A macled crystal is 

 indeed a crystallographical possibility ; but then the axes of the two 

 individuals cohering by the macle-plane have to each other a definite 

 geometrical relation in space, as is well exemplified in the case of the 

 interrupting film in Iceland spar. At the temperature at which 

 "regelation" takes place (viz. the precise limit between the liquid 

 and solid states), it seems to me very possible that the cohesive forces 

 of the molecules of the cemented surfaces may be so nearly coun- 

 teracted as to bring those surfaces into what may be so far regarded 

 as a viscous state as to permit (not indeed a sensible and finite change 

 of figure of a small portion of the mass without fracture, but) a cer- 

 tain freedom of movement in the individual molecules, to some sen- 

 sible depth within the surface, so as to allow of a gradually progressive 

 deviation of their axes from exact parallelism, and thus to effect a 

 transition from one crystalline arrangement to another — not by 

 macling, but by curvilinear distortion, such as may be conceived to pre- 

 vail in pearl-spar and other similar disturbed forms of crystals. Nay, 

 I can conceive it possible that by very long continuance at this exact 

 temperature (especially if aided by tremors short of disruption pro- 

 pagated through the mass, which, as we see in the crystallization of 

 cold wrought iron in the axle-trees of railway carriages, powerfully 

 favour the crystalline rearrangement of molecules even in the most 

 rigid solids) the contiguous blocks may influence each other's crys- 

 tallization to a greater and greater depth through the medium of the 

 cementing film, thus tending continually to straiten the curve of the 

 connecting chain of axes, and after a very long time to bring the 

 two blocks into perfect conformity, so as to form an uninterrupted 



