1920.] Hardcastle.—Deposition of Ice by a Glacier Spring. 27 
the left in the picture). This “ spout ” was about 4 ft. wide and probably 
10 in. or 12 in. deep, and the issuing water made a fine turmoil on striking 
some big stones just outside the area shown in the photograph. 
I saw that the wall was built up of small crystals mingled with some 
sand. A few small masses of crystals came up in the spring. I caught a 
few of these in an eddy pool just below, the outrush, and found them to be 
loosely aggregated granular crystals resembling a grocer’s somewhat coarse 
sugar. They were so loosely aggregated as to be squeezable like a very soft 
sponge. The quantity of ice issuing in the water from the pool appeared to 
be small. On a second visit, a few days later, I used a small improvised 
landing-net made of butter-muslin, with a mouth of about 20 square 
inches. By submerging this in one of the small spout streams enough 
crystals were caught in about a quarter of an hour to make a good-sized 
snowball. A large quantity of water passed through the net in that time. 
On my second visit to the spring I found that considerable changes 
had taken place, but being without the camera I obtained no pictorial 
record of them. The surface of the pool was a few inches lower, and dis¬ 
closed a narrow shelf of solid ice between the spring and the moraine-covered 
mounds. The frontal spout previously seen had closed, and the main dis¬ 
charge was now on the farther side (the right, looking down-stream). What 
had previously been the internal partition seen in the photograph was now 
the outer wall of the spring on that side, and the old outer wall had been 
removed, except a few feet near the left (northern) bank, where it was 
somewhat or wholly shaded from the sun. As before, the water lapped 
over the wall everywhere, with many small spout escapes besides the 
principal one. The area of the pool was reduced by the change, but not 
the area of the spring ; and the spring was as quiet as before. 
The remaining part of the old wall was dry and accessible, and we 
chopped and hoed it with an ice-axe, and found it to be composed of the 
same sort of crystals that we caught in the net, mixed with sand enough 
to make it grey. It was of the consistency, say, of garden-soil that has 
not been disturbed for a year. A part of the wall round the pool was well 
seen, and this was of precisely the same nature, but showed a few streaks 
more sandy than the rest. It was about 3J ft. high, as seen, and probably 
did not descend much below what was visible. The wall was compact 
enough to prevent water from oozing through it. A night’s rain occurred 
between the two visits, and this possibly had something to do with the 
lowering of the height of the wall by increasing somewhat the temperature 
of the water in the pool, for a time. Such a wall must be a temporary 
thing, as its formation and maintenance depend upon dry weather, or at 
all events upon little rainfall. A heavy rain, by diluting and warming the 
seepage from the glacier, would put a stop to the formation of ice-crystals 
and cause a rapid melting of the encircling wall. 
The top of the wall visibly sloped inward. The muddiness of the water 
prevented its being traced more than 1 ft. or so, but the wall could not have 
been more than 3 ft. thick, as the water welled up at that distance from 
the outside of the wall. The rising pipe would seem to have been irregular 
in some way, as the water rose at several points, as it were, alternately, but 
often at two or three points at once. The fact that the water welled up 
quietly and caused no swelling of the surface gives no indication of the 
“ head ” of the subglacial or englacial stream, as the channel may have been 
constricted and the stream “ wiredrawn ” somewhere in its course. 
The walling-in of the spring by the deposition of ice-crystals results 
from the law that pressure lowers the freezing-point of water. Every part 
