340 Transactions. 



north, are destined to obliteration by the rapid lowering of their outlets 

 and by filling up by detritus poured in by waste-laden streams. There 

 appears to be no reasonable doubt that in these cases, too, water has col- 

 lected in the hollows excavated by former glaciers. 



The only other hypothesis which may be put forward to explain these 

 rock-bound basins is the somewhat unsatisfactory one of faulting or warping. 

 Although a system of radiating faults has been suggested as the reason 

 for the peculiar orientation of the valleys of Canterbury (McKay, Geological 

 Report, 1892), the basins, as they occur, could only be explained by a series 

 of peripheral faults or lines of warping disposed in a rude circular ar- 

 rangement around a centre. Of this there is no evidence at present, and, 

 unless the occurrence of earth-movements such as these can be thoroughly 

 demonstrated, it seems best to adhere for the present to glacier excavation 

 as the most satisfactory explanation for the formation of these lake-basins. 

 Further, if earth-movements are a prime cause of their formation, why are 

 lakes found only in those parts of the South Island which have till recently 

 been subjected to the action of ice, and not found in parts which have 

 undoubtedly experienced very recent faulting and other dislocation ? 



In the face of this contention the absence of waterfalls and hanging 

 valleys and the accordance of the grade of the tributaries with that of the 

 main stream may be accounted for in two ways : — 



(1.) That the solid floors are really discordant, but the valley of the main 

 stream and the lower part of the tributary have been so filled with detritus, 

 that the discordance is completely masked. We have really no idea of the 

 depth to which these valleys have been filled with waste, but it may amount 

 to hundreds of feet. Borings in the bed of the Waimakariri in a similarly 

 situated position disclosed nothing but shingle for 30 ft. This is a very 

 shallow depth, but it represents all that has been done in the way of explora- 

 tion by boring in these river-beds. If the thickness be very great, as it 

 probably is, then the tributary valleys of the Rakaia may resemble, perhaps 

 remotely, the tributary valleys of the Milford Sound, such as Sinbad Valley 

 and Harrison's Cove, which are accordant with the level of the sea, or 

 nearly so, but markedly discordant with the floor of the sound. 



(2.) It is possible from the arrangement of the tributary valleys that 

 they were filled with ice long after it retreated from the main valley. The 

 latter for some distance makes a very small angle with the direction of the 

 main range, where the snow collects and forms glaciers. At the present 

 time the Ramsay Glacier runs at right angles to the main valley, and ends 

 on reaching it. The same would be true for every tributary on the northern 

 side of the river at some previous time. At one stage in the retreat of the 

 ice a great river would run across the terminal faces of several large glaciers 

 coming in from the north and occupying a series of parallel valleys. Ice 

 erosion would therefore proceed in them after it had ceased in the main 

 valley. The same accordance in the tributaries can be observed in the 

 Waimakariri, and the same explanation fits this case as well. However,, 

 there are facts which undoubtedly lean to the other side. The valley system 

 had no doubt reached a mature stage, and the valleys were accordant before 

 they were modified by glacier-action, and the fact that they are still ac- 

 cordant may be taken as proof that glaciers have little power of differential 

 erosion. If it were not for the evidence that they have eroded their beds- 

 lower down the valley, I should be inclined to say that the advantage lay 

 with the opponents of ice erosion. My opinion is, however, greatly influenced 



