414 



NATURE 



[September i, 1898 



country. Dr. Nansen, however, might reply that, after 

 all, the existence of this mountain-range is problematical, 

 and that neither in Scandinavia nor the British Islands 

 did the ice-shed and the height of land coincide. Thus, 

 in the north of Ireland the ice-shed of Pleistocene times 

 lay over the central low grounds, while in the north-west 

 of Scotland it occurred east of the water-shed, and 

 the same in a more marked degree was the case in 

 Scandinavia. 



Turning to the much-discussed subject of glacier 

 motion, we find that Dr. Drygalski comes to the con- 

 clusion that movement is the result of variations in the 

 mass of the ice. Numerous observations and measure- 

 ments demonstrated that there is both a vertical and a 

 horizontal movement in the " inland ice," the former being 

 the primary jnovement of the two. Over the marginal 

 zone he observed a well-marked bulging of the surface, 

 while further inland, where the ice is thicker, the surface 

 appears relatively depressed — a condition sometimes ob- 

 scured, however, by the heaping-up of snow. These 

 . differences in the configuration of the ice-sheet are due 

 to variations of mass within the ice, the sinking or de- 

 pression being the result of internal shrinkage, which is 

 always greatest at the bottom, and progressively di- 

 minishes upwards. Had the whole mass shrunk in the 

 same proportion as the ice at the bottom, the sinking at 

 the surface would have been more pronounced. 



The stratified or bedded structure of the ice has the 

 same tale to tell. That structure is the result of the 

 freezing of water under pressure, and since the individual 

 layers diminish in thickness from below upwards, while 

 the cold at the same time increases, it is clear that the 

 internal shrinkage under which refreezing takes place 

 must likewise lessen towards the surface. It is evident, 

 indeed, that the layers must become thinner upwards, 

 seeing that the pressure necessary for their formation 

 diminishes in that direction. Melting, no doubt, does 

 take place at the surface, and the released water trickling 

 downwards is again frozen, but stratification does not 

 result from this process. It is at lower levels in the ice 

 that the structure is developed. And as water cannot 

 possibly filter down from the surface through a compact 

 ice-mass, the obvious conclusion is that the water neces- 

 sary for the production of the structure in question 

 originates within the " inland ice " as the result of pres- 

 sure. The presence of stratification, then, shows that 

 liquefaction and re-solidification take place in the "in- 

 land ice." But the water set free under pressure cannot, 

 as a rule, refreeze in exactly the same place, otherwise 

 it would be difficult to account for vertical movement in 

 the ice. 



Depression of the surface indicates a diminution, and 

 bulging of the surface an increase in the volume of the 

 ice. Under the weight of the overlying mass material 

 is squeezed out from the thicker into the adjacent thinner 

 portion of the ice. In short, an outflow takes place, and 

 will continue as long as a sufficient degree of melting 

 is kept up in the former, and the same degree of mo- 

 bility is not attained in the latter. The ice-sheet, there- 

 fore, moves from the interior, where it is thickest, to the 

 marginal area, where it is thinnest. And observation 

 showed that under these conditions it could move up 

 slopes. 



Dr. Drygalski points out that many complications 

 arise from the varying distribution of heat in the ice- 

 masses, and from other causes which need not be re- 

 ferred to here. He found that the temperature of the 

 thinner ice of the marginal area was generally lower than 

 that of the thicker ice stretching inland. In the latter 

 the ice is at, or nearly at, the melting-point. There is 

 thus again a tendency to movement from the interior 

 outwards. Water is forced from the thicker into the 

 thinner masses, but, because of the low temperature of 

 the latter, it quickly freezes, and thus gives rise to the 



NO. 1505, VOL, 58] 



formation of new ice-layers. The abundant presence of 

 stratification in the thinner ice of the marginal area 

 shows that this process is very active there, while the 

 bulging of the surface proves that the bedded structure 

 is intimately connected with increase of volume. 



Sometimes the horizontal movement is so pronounced 

 as to obscure the vertical movement more or less com- J 

 pletely. In other places only the latter may be notice- | 

 able. The rate of the former depends on the thickness ' 

 of the, ice and the intensity of the vertical move- 

 ment. The greater these are the more rapid it becomes. 

 In the independent glaciers of the coastal tracts it was 

 found that the rate of motion diminished as the rock 

 debris included in the ice increased in quantity. This 

 was to have been expected, since the mass of the ice, 

 and therefore the whole thickness of the glacier, dimin 

 ished at the same time. In the longitudinal section of 

 such a glacier the rate of motion lessens towards the 

 end, but with the " inland ice " the reverse is the case — 

 it increases. In the former the ice loses bulk absolutely 

 owing to ablation at the surface, and relatively because 

 of the inclusion of rock-rubbish. But the great ice- 

 streams that flow from the interior into the deep fiords 

 increase in thickness towards the end. In glaciers and 

 " inland ice" alike the horizontal movement of the sur- 

 face depends upon that of the lowest layers. At Asakak, 

 for example, the horizontal movement at the bottom was 

 measured and compared with that of the surface, and 

 this proved to be less than it ought to have been if all 

 the layers of like thickness between the bottom and the 

 surface had been moving at the same rate. The differ- 

 ential movement of the individual layers, therefore, de- 

 creases from below upwards. 



The movement at the surface of a great ice-stream 

 coming from the "inland ice" increases towards the 

 end. Were it not for the rapid movement of its lower 

 layers, therefore, the ice-flow would lose its continuity. 

 When the ice enters the sea, it eventually reaches a 

 point where the pressure of the mass itself no longer 

 affects the lower layers— the primary vertical and secondary 

 movements cease, the squeezmg-out process comes to an 

 end, and true glacier motion is succeeded by the purely 

 passive movement of the iceberg. 



In his discussion of the mechanics of glacier motion, 

 Dr. Drygalski, as will be seen, upholds the well-known 

 theory of Prof. James Thomson. He points out how the 

 water set free under pressure is transfused into air 

 vesicles, cracks, &c., in the ice, where it freezes again, so 

 that the ice eventually becomes clearer. As this process 

 goes on most rapidly at the greatest depths, the ice at 

 the bottom is necessarily the clearest— clearness, in 

 short, increases from the surface downwards. Further, 

 since refreezing takes place under pressure, the ice 

 crystals arrange themselves with their chief optic axes 

 perpendicular to the lamination or bedding of the ice. 

 As a result of these changes, the volume of the ice 

 is diminished — the shrinkage being greater in the 

 thick than the thin layers, and more marked in the 

 inland tracts than in the marginal area of the ice- 

 sheet. But we need not follow the author further 

 into this part of his discussion. When he states that 

 the horizontal motion depends upon the movement 

 of water within the ice, he will not be misunderstood. 

 He does not mean free flowing streams of water, but 

 mechanical changes in the mass a"nd transference of 

 conditions. Perhaps also it may be as well to add that, 

 although measurements prove that differential move- 

 ment of the ice-layers increases from the surface to the 

 bottom, it is not to be supposed that one layer flows 

 out from under the layer above it. There is a certain 

 loosening of the connection between them, the author 

 remarks, but not an actual separation. In consequence 

 of this some of the motion of the lowest layer is added 

 to that of the next above, the rate of which is thereby 



