202 MR. E. J. GARWOOD & DR. J. W. GREGORY [May 1 898, 



But the most important difference between Arctic and Alpine 

 glaciers is seen in the character of their respective terminal fronts. 

 Most Swiss glaciers end with a tapering snout. Some of those in 

 Spitsbergen (such as Flower Glacier, Baldhead Glacier, etc.) do the 

 same ; but the majority end in a vertical cliff of ice, of the type 

 for which Lieut. Lockwood's apt name of ' Chinese wall ' ^ is gene- 

 rally adopted. This name is appropriate owing to two characters : 

 the face is vertical, and sometimes overhangs at the top like the 

 machicolations of a mediaeval fortress ; secondly, the intraglacial 

 material is arranged in lines like the layers of mortar in a wall. 



The vertical faces of these Arctic glaciers have been suggested 

 by Chamberlin^ as largely due to the low angle at which the sun 

 shines upon them. This no doubt assists in the formation of these 

 Chinese walls,' for when the sun is low on the horizon it tends 

 to cut the opposite face of the glacier backward by vertical ablation, 

 instead of reducing its thickness by ablation over the whole upper 

 surface. But the position of the sun is not the only factor, and we 

 doubt whether it is a very important one, for the raj-s of the sun 

 when it is near the horizon have comparatively little melting-power, 

 compared with those which come from the zenith. Moreover, 

 glaciers facing in the same direction and in the same neighbourhood 

 sometimes end with tapering snouts, and sometimes with Chinese 

 walls. (See PL XVIT, fig. 1.) 



Our first impression on seeing a glacier with a wall-termination 

 was that this arrangement was due to the glacier being in a state 

 of advance : and, so far as we could see, glaciers ending in snouts 

 were either receding or stationary, while those with Chinese walls 

 were always advancing. 



The formation of the overhanging cornice appears to be due to 

 the rapid forward movement of the uppermost layers of ice. This 

 may be aided to some extent by the melting backward of the lower, 

 dirt-bearing layers, owing to their greater absorption of heat ; but in 

 the case of at least some of the Spitsbergen glaciers, the influence of 

 this factor must be quite insignificant.^ 



Glaciers with these Chinese walls are more instructive than those 

 with snouts, and we therefore propose to describe one in some 

 detail. The first view of such a glacier (see PI. XIIT, fig. 1) shows 

 that it is composed of a lower part charged with debris, and an 

 upper part in which the ice is pure and white. We almost 

 instinctively proceeded at once to the foot of the ice-cliff to search 

 for ' ground-moraine.' But the included material in the lower 



^ A. W. Greeley, ' Three Years of Arctic Service,' vol. ii (1886) p. 34, & 

 app. p. 28, figs. 3 & 4. 



2 T. C. Ohamberlin, ' Glacial Studies in Greenland,' pt. vi, Journ. Geol. 

 vol. iii (1895) p. 566. 



^ Its action has been invoked to help to explain vertical glacier-faces by 

 E.eid, and overhanging faces by Salisbury and Chamberlin. See H. F. Beid, 

 ' The Mechanics of Glaciers,' Journ. Geol. vol. iv (1896) p. 926 ; R. D. Salisbury, 

 ' Salient Points concerning the Glacial Geology of North Greenland,' Journ, 

 Geol. vol. iv (1896) p. 782 ; and T. C. Chamberlin, ' Glacial Studies in Green- 

 land,' pt. ix, Journ. Geol. vol. iv (1896) p. 591. 



