112 R. MW. Deeley—Striations on Glacier Granules. 
is smaller than the hind; the humerus is only slightly expanded at 
its lower end, and the rest of the paddle is long and slender. The 
pelvis consists of small rod-lke, backwardly sloping ilia, great thin 
plate-like pubes, and the greatly elongated ischia characteristic of the 
family. As usual in the group, the ium is not in contact with the 
pubes in the acetabulum. The greatly expanded coracoids, pubes, and 
ischia formed an almost continuous bony floor to the body, and the 
short interval between the posterior edge of the coracoids and the 
anterior edge of the pubes was filled by a plastron of ventral ribs; in 
the mounted. specimen this is represented only by three of the median 
ventral ribs, which are fused with one another on the middle line. 
The hind paddles, though larger, are closely similar in form to the 
_ fore paddles. The total length of the specimen as mounted is 11 ft. 6 in. 
The dimensions in centimetres of some parts of this skeleton are— 
Skull, length . : per 8, 3 : 55°7 centimetres. 
Mandible, “leneth ; ‘ : ; ; 67°0 Bi 
as length of symphysis 21:4 5 
Coracoid, greatest length : f : 47:0 5 
Humerus, ‘length : : : : : ; 33°0 - 
Pubes, leneth 3058 i 
Ischium, length of median expansion 37:0 + 
Femur, leneth 39°0 
39 
V.—Gatacrer GRANULE-MARKINGS. 
By R. M: Drevey, M.Inst.C.E., F.G.S. 
(PLATE XIII.) 
HAVE already described the granular appearance of glacier ice 
as seen in polarized ight! and also the striations on the granules 
as shown by pencil rubbing gs.” Last summer I succeeded in obtaining 
exact reproductions, in plaster of Paris, of the ice surface-structure in 
the upper cave in the Rhone Glacier. These are shown in Figs. 1-8. 
The casts were obtained in the following manner. Plasticene, 
a substance used for modelling, having been cooled to the temperature 
of the ice, was pressed against the wall of the cave. When all the 
conditions were favourable the surface of the plasticene took an exact 
impression of the ice-surface. A cardboard ring, obtained froma pillbox, 
the top and bottom of which had been removed, was then pressed 
upon the surface of the plasticene and filled with liquid plaster of 
Paris. When the plaster had properly set the plasticene and card- 
board were removed. Photographs of these casts, as will be seen 
from the photos reproduced, give very good representations of the 
ice markings (see Plate XIII). 
A careful examination of well-marked granules in which lquid 
cavities had been produced by a burning-glass showed that the ridges 
and furrows in each grain were along planes at right angles to the 
optic axes of the granules. It is clear, therefore, that the direction 
of the markings is determined by the crystalline structure of the ice. 
Fig. 2 is a print of a portion of the surface of a large glacier 
eranule, and Fig. 1 shows portions of several granules. ‘Fig. 3 is 
1 Grox. Mac., 1895p. 152. . 2 Thid., 1907, p. 529. 
