GLACIOLOGY OF THE SOUTH ORKNEYS. 839 



miles even at the heads of bays, where the ice-sheets attain their fullest development. 

 Along the sides of the bays the length of the ice-sheets, i.e. the length in the direction 

 of flow, may be taken as being more commonly under than over half a mile. Their 

 width, however, is often considerably greater than their length. Since the ice but 

 seldom occurs in the shape which we associate with the classical description of a glacier 

 (glaciers properly so-called, such as Alpine valley glaciers), I have used, as a rule, the 

 term " ice-sheet " as being more correctly descriptive, although the term glacier has 

 occasionally been used as an equivalent. Goukdon's term of "piedmont" glacier is 

 literally accurate in describing these ice-sheets, but it introduces confusion by using a 

 term which has already been ear-marked as descriptive of glaciers which are really of 

 quite a different type. Nordenskjold's term of ice-foot glacier is probably ^ on the 

 whole, the best title for this particular type of ice-sheet, as it emphasises the resemblance 

 between them and the snow formation which, although at a lower level and on a much 

 smaller scale, also forms a terrace round Antarctic lands, viz. the ice-foot (see p. 848). 



Details of Structure. 



These ice-sheets may be divided into three parts — the initial steep portion on the 

 upper hill slopes ; the main part, more nearly level ; and the terminal ice-cliff or snout, 

 as the case may be. 



High Slopes. — -The first portion is very steep, and may extend right up to the 

 summit of the hills or dividing ridge, but more commonly it commences some little 

 way below the top ; it may have a slope of as much as 25° or 30°. It is sometimes 

 marked off from the second stage by a well-defined " Bergschrund," but whether this is 

 commonly the case or not I do not know, as the surface snow-covering was usually so 

 perfect that all crevasses were bridged over. In the instances where one was noted it 

 was never of any great width (see Plate IV. fig. 1). 



Snow-caps. — Snow-caps accumulate on the tops of some of the hills either merging 

 into the initial high slopes or forming cut-off snow-caps above the generally glaciated 

 level. This material is dissipated to some extent by melting and ordinary flow, but is 

 largely got rid of in the form of avalanches on to the more level portions of the ice- 

 sheet below (see Plate VII. fig. 1 ). I examined several blocks that had fallen from such 

 positions, and found them to have a well-marked stratified structure, formed of alternate 

 layers a few inches thick of blue transparent ice and white opaque n^ve-ice resembling 

 exactly the uppermost layers of the glaciers themselves, as described later (see p. 856). 



Like the glaciers in Spitsbergen described by Garwood and Gregory (16), these 

 ice-foot glaciers therefore do not drain snow-fields ; the snow passes directly into the 

 condition of nev6-ice and glacier-ice. This can be seen very well in the crevasses in a 

 col. Looking down one of these there is seen first a layer several feet thick of 

 alternating thin bands of blue ice and white neve-ice, the latter gradually become 

 thinner and thinner until, a very little way below the surface, there is nothing but 

 layers of ice, still, however, containing a large proportion of air-bubbles and not of the 



TRANS. ROY. SOC. EDIN., VOL. XLIX., PART lY. (NO. 15). 114 



