BY PROFESSOR GRIFFITH TAYLOR, D.SC, B E., B.A., F.G.S. 197 



There are two possible explanations of this unusual shelf 

 with its rock-tarns. One involves the filling of the whole 

 Broad Valley with ice, so that the lateral drainage flowed 

 to the north along the position of this shelf and so cut a 

 notch betv/een the glacier and the containing ridge to the 

 west. This is not well supported by the field evidence, though 

 it accounts for the shelf sloping to the north. 



The more plausible explanation involves the nivation- 

 layer which I have described above. I imagine that for some 

 long period this layer with a temperature around 34 deg. 

 F. halted at the 4,300 feet level, possibly both in the advanc- 

 ing and retreating hemi-cycles of the Ice Age. (See Fig. 4 

 above.) The shelf was favourably situated for collecting 

 snow, which was not readily removed by the sun from its 

 sheltered position. A series of cirques were sapped out in 

 the course of time, and these became apposed sideways in 

 much the same fashion as Nussbaum has described in the 

 Swiss Alps. (6) 



A shelf is thus produced by the sapping action of seven 

 adjacent cirques. The ice-slabs are competent to carry the 

 erratics to the positions noted, and also to round the rocks < 

 forming the rim of the shelf. Since a cirque glacier ''btir- 

 roivs" into the hill (as Hobbs has shown in his "Characteris- 

 tics of Existing Glaciers") rather than erodes the valley 

 under its snout, we see why the edge of the shelf remains 

 almost entirely unaffected by the shelf glaciers. 



The evolution of Lake Belton, perched some 300 feet 

 up the side of the Lake Belcher Valley, may be partly ex- 

 plained in' a similar fashion, but this demands much more 

 field work than I was able to give to this locality. 



It is in the hope that these brief notes will stimulate local 

 interest in the innumerable geographical problems of Tas- 

 mania that I have written the paper. 



LIST OF ILLUSTRATIONS. 

 Plate XXIX. 



Sketch survey of the Park. The form-lines are approxi- 

 mately correct near the routes marked, but are only filled 

 in from sketches, etc., elsewhere. 



(6) Die Taler der Schweizer-Alpen ; Berne, 1910. 



