80 Rev. E. Hill — Questions and Answers on Ice-Motion. 



shape ? And again, liow can the so-called Glacier-tables subsist ? 

 Why does not the ice of the stem squeeze out under the weight of 

 the top stone, and let it down ? 



I do not write for mathematicians and physicists, who will smile 

 at these questions, and may improve my solutions. I write for all 

 who having read them must hesitate before answering. These 

 would be ray answers if asked the questions in an examination. 

 Reversing their order : A Glacier-table is due to a melting away of 

 the exposed surface by the sun, while the ice beneath a large stone 

 is unmelted. If the surface sinks by that waste faster than the 

 protected stem shortens by compression, the stem will lengthen. 

 The squeeze might make the stem thicken, and should, unless the 

 waste of thickness by evaporation, etc., exceed the increase which 

 the compression can produce. Crevasses are due to a tension acting 

 on the mass of ice. If viscosity cannot close the fissure faster than 

 tension is opening it the crevasse once open must remain so. When 

 the flow of the glacier has carried the ice past the tension-position 

 the crevasses will close. The Antarctic ice, if a viscous body, must 

 flow outwards into the ocean until it is ready to float, and till its 

 strains from floatation or other causes are more than its strength can 

 bear. At this- point masses will break off and float away, leaving 

 where they broke from the wall-like face described. A similar wall 

 on a small scale due to a like cause bounds (or did, when I saw it, 

 bound) the Aletsch glacier along the Marjelen See. 



As another question : If that ice be viscous, why does it not flow 

 up the Marjelen side-valley as well as down the main Aletsch 

 channel? Answer: It does so flow, to an extent measured by the 

 amounts melted away from that face, and broken away. 



The following, less simple, problem may enable some readers to 

 test their ideas on the matter of ice moving *• uphill." I suppose 

 everyone has seen the undulated erection called a switch-back 

 railway. Suppose a continuous train of cars in contact along the 

 whole length of rails : if free to move, the weight of the topmost 

 ones will force the whole train forward. Suppose the cars replaced 

 by blocks of ice, separate but in contact ; no doubt a similar motion 

 will ensue. But now, suppose the interstices between successive 

 blocks all filled up, so that they are united into one continuous thick 

 ribbon, or undulated beam of ice ; what will happen then ? 



Then, in motion forwards, convex parts would have to compress, 

 concave parts to extend. I have no doubt that if the upper end be 

 at a level high enough, some motion will result. "What that motion 

 •will be, whether the same forward flow as of the separate blocks, 

 or a shattering to pieces of the ice at points of greatest stress, or 

 a thickening in the hollows, or a combination of these, depends on 

 the amount of breaking-strain and the rate of viscous flow. If, 

 however, the stress be enough to cause any viscous yielding, the ice 

 in the hollows will thicken, the mass forced uphill will increase, 

 the force required to move it will in turn increase, and the thickening 

 be accelerated (assuming, of course, a continual supply of ice at the 

 starting-point); whilst, on the other hand, it may be seen that con- 



