March 24, 1870] 
IATROTAS 
539 
brought into contact with a similar surface, it unites with 
it so as to form one continuous mass. Between the resist- 
ance to shearing and the force which tends to bring the 
glacier down there must be a mechanical relation, so that 
if the shearing resistance were greater, the force would be 
insufficient to cause the descent. By a series of experi- 
ments upon cylinders of ice inserted in a cylindrical hole 
bored through two pieces of wood perpendicularly to the 
surface along which the*one was made to slide upon the 
other, it was found that the force necessary to part the 
ice along the sliding surface varied from 75 to 119 lbs. 
per square inch. Canon Moseley has calculated that for 
the Mer de Glace to descend by its own weight, its shear 
per square inch cannot exceed 1°3193 lbs., and that to 
produce the actual motion with a shear of 75 lbs. per 
square inch, a force in aid of the weight and thirty-four 
times as great must be called into existence, and applied 
in the direction of motion. For such a force to be pro- 
duced by the weight of the glacier alone the density of 
ice would require to be increased more than 400 times. 
In this reasoning Canon Moseley has neglected, as it 
appears to ine, the capability of ice when in a state of 
deliquescence to slide along a surface of small inclination, 
as demonstrated by the well-known experiment of Wil- 
liam Hopkins. It is, however, not the motion of a block 
of ice as a whole, but the differential motions of 
its particles that we have now to consider. It occurred 
to me that the Canon’s arguments upon this branch of 
the question might be put to an easy practical test by 
subjecting a block of ice to a strain produced by its own 
gravitation, and observing its behaviour under this condi- 
tion, and I was fortunate in obtaining the assistance of 
my friend Mr, A. F. Osler, F.R.S., in carrying out the 
experiment. 
A plank of ice 6 inches in width and 22 inches 
in thickness was sawn from the frozen surface of a 
pond, and supported at each end by bearers exactly six 
feet apart. From the moment it was placed in position 
it began to sink and continued to do so until it touched 
the surface over which it was supported, drawing the 
bearers with it, so as to make their upper ends converge. 
At its lowest point it appeared bent at a sharp angle, 
and it was rigid in its altered form. The total deflec- 
tion was 7 inches, which’ had been effected in about as 
many hours under the influence of a thaw, during which 
the plank diminished slightly in width and thickness. 
On observing the under surface of the plank near the 
point of flexure, I noticed a number of very minute fis- 
sures extending a short distance into the ice, but they cer- 
tainly were not sufficient to account for the flexure of the 
plank. 
The question at once suggested itself, was the change 
of form in the ice plank due to fracture and regelation? 
I did not think it was, but the experiment was not deci- 
sive. Some weeks afterwards an opportunity occurred of 
trying it under other conditions. During the last frost 
we cut out another ice-plank. Its length was 6 feet 9} 
inches, its width varied from 6} to 6} inches, and its 
thickness was If inches. Two large bricks, of a width 
exceeding that of the plank, were set up on end, on a 
horizontal surface, exactly 6 feet apart, and the plank was 
laid upon them at five p.m. on the 12th of February. At 
3/15 p.m. on the 13th it was continuously curved from end 
to end, so that it only rested on the edges of the bearers, 
and the middle point of its upper surface was deflected 
1? inches below the line joining its two extremities. The 
temperature was 26°F. The curved plank was perfectly 
rigid, as was proved by taking it off the bearers and in- 
verting it. I examined it again on the two subsequent 
days with the following results ;— 
Feb. 14th, 9.30 A.M. Temp. 29°5 F. 
Deflection of upper surface below chord . 2% inches 
- of lower surface below its ori- 
ginal horizontal position . . 2} ,, 
Feb, 15th, 9.30 A.M. Temp. 30° 0 F. 
Deflection of upper surface below chord 3$ inches 
7] of lower surface below its ori- 
ginal horizontal position . . 3 
Cae 
” 
During the whole of this interval, in which the tempera- 
ture never rose above the freezing point, there was 
no indication of fracture in the plank, nor did the optical 
continuity of the ice suffer the slightest interruption. 
On the 15th it began to thaw, and the bearers having be- 
come frozen to the ground, and the plank to the bearers, 
the suspended portion was unable to yield to the strain 
produced by its gravitation ; and when I re-visited the 
plank on the afternoon of the 15th, it was broken into 
half-a-dozen pieces. 
These experiments were very rough and imperfect ; we 
intend to 1enew them on some future occasion, and to 
conduct them with much greater care and proper me- 
chanical appliances, when we hope to be able to bend 
an ice-plank double, without destroying its continuity. 
Ae following conclusions may fairly be drawn from 
them :— 
1.—A mass of ice may change its form under strains 
produced by the gravitation of its particles, without 
becoming fractured, and without returning to its 
original form when the strain ceases. 
2.—The change of form takes place at tempera- 
tures both below and above the freezing point, but 
is greatly accelerated in the latter case. 
I shall not now attempt to discuss the nature of the 
molecular displacements to which the change of form is 
due. Their occurrence is indisputable ; whether or not 
they are to be dignified by the name of shearing is a mere 
verbal question of little moment. In a very able paper in 
the PAzlosophical Magazine for March 1869, Mr. James 
Croll adduces good reasons for believing that when a 
mass of ice has a deliquescent surface, its molecules may 
experience repeated momentary losses of their shearing 
force. While, therefore, he admits the conclusiveness of 
Canon Moseley’s reasonings for temperatures below freez- 
ing, he conceives that ice at all higher temperatures may 
shear by its own gravitation. It is evident that the former 
concession in Canon Moseley’s favour cannot now be 
maintained, and that the point to which our experimental 
researches should be directed is not what amount of force 
will suddenly rend asunder the molecules of ice beyond 
the sphere of their mutual attractions, but what amount of 
force will produce molecular displacement within that 
sphere, with time allowed for its operation. 
If we conceive an ice-plank, instead of being placed 
horizontally between bearers, to be laid with its narrowest 
face upon a plane of small inclination, with its upper edge 
horizontal, and its ends confined between vertical walls 
converging in the direction of motion, with its under sur- 
face deliquescent, so that friction would almost be annihi- 
lated; and if we further imagine the diminution of gravity 
due to resolution along the plane to be compensated 
by increasing the length or diminishing the thickness of 
the plank, the plank would alter its form in a way pre- 
senting a striking resemblance to the actual movement of 
a glacier. Its central portions would move more rapidly 
than its lateral ones; its surface more rapidly than 
its base; and when the strain upon its particles ex- 
ceeded their cohesive power, it would fracture obliquely to 
the axis of the channel. 
If the conclusions drawn from the experiments above 
described are legitimate, plasticity must be admitted by 
the side of sliding, and fracture and regelation as one of the 
constituent elements of the theory of glacier motion, and 
a more important place in that theory must be assigned 
to the views of the late Principal Forbes than has for 
some years been conceded to them. 
WM. MATHEWS, Jun. 
