Sir IT. n. Hoivorth— Erratic Boulders in Drift. 127 



the solid materials it comes across, and what it would have to do 

 to overcome gravity in this way. 



It is quite true that stones sometimes occur imbedded in the mass 

 of a glacier, and afterwards appear at its surface, and thus super- 

 ficially seem as if they had travelled upwards ; but this is due to 

 a very different process. If a stone falls down a crevasse, and is 

 arrested before it reaches the floor of the glacier, it will, wben the 

 crevasse closes up on level ground, be imbedded in the ice. If, again, 

 a stone or a group of stones lie on the back of a glacier, and a suc- 

 cession of snowfalls come upon it, which snow is converted into ice 

 subsequently, it of they will become imbedded in solid ice, and in tbis 

 way so-called englacial drift ma}' be formed. When by ablation or by 

 the direct rays of the sun a certain quantity of the surface of the 

 glacier is subsequently removed, these imbedded stones will appear, 

 like tbe imbedded remains of the famous Austrian travellers lost 

 three centuries before they appeared on the surface of the Glacier of 

 Tliisdiil in 1885 ; but this is not by any process of travelling 

 upwards through the ice from its bed to its surface, for which no 

 mechanical explanation seems in any way available, but by the un- 

 covering of what was always in mid-ice. What is true of travelling 

 upwards is equally true of travelling downwards, or perhaps more 

 true, for it involves the stones travelling from less to more dense 

 strata of ice, and from layers under slight pressure, and therefore 

 friction, to layers under great pressure and great friction. The 

 stones will, in fact, move with the la3'ers in which they are im- 

 bedded, and when a glacier is moving on n rapid slope, as Spencer 

 noticed by going underneath such a glacier in Norway, the scattered 

 stones, even at its base, will he sloivlij moved — very slowly, since he 

 noticed that the ice itself, which at this level crept very quietly, and 

 the layers above, were apt to move over the stones and to have 

 gashes cut into themselves by them. This was on a sleep slope. On 

 level ground, as I have said, the capacity of the nether layers of nny 

 ice mass which we can conceive of, for any appreciable motion or fur 

 exercising any appreciable thrust upon the stones beneath it, is most 

 doubtful. 



But let us pass on, and suppose it had the capacity. What must 

 happen ? A stone falls to the bottom of a crevasse. The ice begins to 

 move, and the stone is rapidly enveloped in ice, and is held in its grasp 

 as in a vice. As the ice moves on and exercises a certain thrust upon 

 the stone, it will act like the slipper or skid of a coach, and its lower 

 surface in contact with the glacier's bed will be worn down and 

 smoothed and polished. If presently the stone be arrested by some 

 obstacle, and the ice flows over it, it will in some measure also rub 

 down the rugosities of the upper surface, but probably not to the 

 same extent. The stone will thus have two more or less parallel 

 faces — one a good deal smoothed, and the other not quite so much — 

 but the other angles or edges of the stone will remain intact, and 

 unrounded and unsmoothed. We cannot understand how by any 

 process a glacier can make a true boulder with subangular edges or 

 a curved outline all round, or in the case of soft polygonal stones 



