216 THE JOURNAL OF GEOLOGY. 
ends or sides of the glacier from which they plunged downwards 
in beautiful little cascades. 
Perhaps the most interesting feature of the surface of the gla- 
cier was its numerous dust wells, a phenomenon which Norden- 
skjéld brought pointedly to public attention some years since. 
Upon this glacier they were relatively small but exceedingly 
numerous and widely distributed. They are cylindrical tubes 
penetrating the ice to a depth of six or eight inches, or occasion- 
ally a little more. They ranged in size from tubelets which 
would scarcely more than admit a lead pencil up to wells of a 
foot or more in diameter, though the latter sizes were rather 
rare. Occasionally they were double or complex, due perhaps 
in part to the joining of two or more smaller tubes, and in part 
to original irregularity in the distribution of the dust which formed 
them. Sometimes there was a central column of ice reaching 
from the bottom to a capping above which implies the former 
explanation. The tubes were singularly vertical. Nowhere were 
they observed to incline to the northward, as holes melted in the 
snow in southern latitudes so commonly do. There was, how- 
ever, no question as to their being the effects of melting. At 
the bottom of each was a thin film of black dust which was their 
obvious cause. This dust, catching the sunlight and transform- 
ing its energy, melted its way downward. The circular course 
of the sun doubtless tended to correct any tendency to inclina- 
tion, but even in this latitude the difference between the heat of 
the southerly and the northerly sun is quite pronounced, and the 
verticality is apparently independent of the inclination of the 
sun’s rays. 
It was noticed that a film of ice frequently stretched across 
the upper part of these wells, and that the water within them 
had shrunk away two or three inches from the film. Occasion- 
ally there was a second film of two or three inches below the 
first, and the water had sometimes shrunk away from this also. 
Obviously the well had been nearly full when the first film of ice 
was formed, for the most of them were near the mouth of the 
well. As these observations were made near midday, the sug- 
