58 
ing opposite angles, speaking roughly, were on an average 
about one inch in Jength; but in another part the forms ap- 
peared rather more regular—hexagons being common—the 
diameters of which were about 4 inch. Another fragment was 
then obtained rather further from the edge; this was a little 
more than one inch thick, and consisted of prisms whose ends 
were about the same size as those just described. Though 
the number of sides in their polygonal ends was not constant, 
six was certainly a common number, and this appeared to be 
rarely exceeded; the angular points of the polygons were a 
little blunted, so that the sides were slightly curved. The 
angles were thus difficult to estimate, and I had no instrument, 
but I do not think that they were constant. The ice broke 
very easily along the sides of these prisms, and never through 
or athwart one of them; so that each fragment had a beauti- 
fully dentate edge (see Fig.), and its side resembled a miniature 
group of basaltiform columns. Single columns could be easily 
detached quite perfect, with the point of a penknife, or even 
with the finger-nail. The ice contained a few air-bubbles and 
a chance bit of weed or bark here and there, but was in no 
other way remarkable. We carried a large piece of this last 
described to my rooms; and after hastily improvising a freezing 
bath of snow and salt, subjected some fragments of it to a 
temperature of about —5° Cent. (for a short time the thermo- 
