622 REPORT— 1904. 
tarns most abundant—that is, where the harder and more compact rocks make the 
work of streams in tapping the basins more lengthy. The rarity of tarns in the 
highlands behind Kanchenjunga, perhaps, calls for explanation. We came upon 
many basins, but, whether formed by moraines or true rock-basins, they had for the 
most part been filled up by alluvial deposits. 
In my opinion, the presence of tarns must be taken as an indication that the 
portion of the range where they are found has until a comparatively recent date 
been under snow or ice. The former theory, still held, was that the ice scooped 
out their basins from the solid rock. I believe that it simply kept scoured pre- 
existing basins. The ice removed and the surrounding slopes left bare, streams 
on the one hand filled the basins with sediment, or, on the other, tapped them by 
cutting clefts in their rims. This theory meets, at any rate, all the facts I have 
observed, and I may point out that the actual process of the destruction of tarns 
by such action may be seen going on under our eyes in many places, notably in 
the glens of the Adamello group. Professor Garwood has lately employed his 
holidays in sounding many of the tarns of the St. Gotthard group, and his results, 
I understand, tend to corroborate the conclusions stated. 
I desire here to reaffirm my conviction that snow and ice in the High Alps are 
conservative agents: that they arrest the natural processes of subaerial denudation ; 
that the scouring work done by a glacier is insignificant compared with the hewing 
and hacking of frost and running water on slopes exposed to the open sky without 
a roof of névé and glacier. 
The contrast between the work of these two agents was forced upon me many 
years ago while looking at the ground from which the Eiger Glacier had then 
recently retreated. The rocks, it is true, had had their angles rubbed off by the 
glacier, but through their midst, cut as by a knife, was the deep slit or gash made 
by the subglacial torrent. There is in the Alps a particular type of gorge, 
found at Rosenlaui, at the Lower Grindelwald Glacier, at the Kirchet above 
Meiringen, and also in the Caucasus, within the curves of old terminal moraines. 
It is obviously due to the action of the subglacial torrent, which cuts deeper and 
deeper while the ice above protects the sides of the cutting from the effects of the 
atmosphere. 
One more note I have to make about glaciers. It has been stated that glaciers 
go on melting in winter. Water, no doubt, flows from under some of them, but 
that is not the same thing. The end of the Rosenlaui Glacier is dry in January ; 
you can jump across the clear stream that flows from the Lower Grindelwald 
glacier, That stream is not meltings, but the issue of a spring which rises under 
the glacier and does not freeze. There is another such stream on the way to the 
Great Scheideck, which remains free when frost has fettered all its neighbours. 
I should like to draw your attention before we leave glaciers to the systematic 
efforts that are being made on the Continent to extend our knowledge of their 
peculiarities. The subject has a literature of its own, and two Societies—one in 
France, one in other countries—have been constituted to promote and systematise 
further investigations, especially with regard to the secular and annual oscillations 
of the ice. These were initiated by the English Alpine Club in 1893, while I was 
its president. Subsequently, through the exertions of the late Marshall Hall, an 
enthusiast on the subject, an International Commission of Glaciers was founded, 
which has been presided over by Dr. Richter, M. Forel, and others; and more 
recently a French Commission has been created with the object of studying in 
detail the glaciers of the French Alps. A number of excellent reports have been 
published, embodying information from all parts of the globe. There has been, 
and is, I regret to say, very great difficulty in obtaining any methodical reports 
from the British possessions oversea. The subject does not commend itself to the 
departmental mind. Let us hope for improvement: I signalise the need for it. 
Of course, it is by no means always an easy matter to get the required measure- 
ments of retreat or advance in the glacial snout, when the glacier is situated in a 
remote and only casually visited region. Still, with good-will more might be 
done than has been. The periods of advance and retreat of glaciers appear to 
correspond to a certain extent throughout the globe. The middle of the last 
