Physiography, Geology.] SUBANTARCTIC ISLANDS OF NEW ZEALAND. 715 



ridges one within the other, showing that glaciation was becoming less intense ; such 

 ridges are frequently seen in those parts of the Southern Alps where glaciers are 

 retreating. 



In addition to this undoubted evidence from the lateral moraines, there are 

 certain collections of rough angular blocks of all sizes which lie across the lower 

 end of several of the tributary valleys, and at times pass into the lateral moraines. 

 These valleys are distinguished by their cross-section. They are usually wide and 

 U-shaped ; they are never narrow ; and the streams which occupy them have no 

 well-defined channel till they cut through the angular debris lying across their mouths. 

 During this portion of their course the streams never expose solid rock, and the 

 boulders are of such a size that they could not owe their present position to the 

 transporting agency of these streams. They are undoubtedly the terminal moraines 

 of small glaciers, and are now so completely covered with peat and scrub that they 

 can only be examined satisfactorily by wading down the beds of the torrents which 

 cut through them. 



The heads of the valleys show the typical amphitheatre form which charac- 

 terizes the upper part of glacier-troughs, so that they were once occupied by ice, 

 and the irregular accumulations lower down the valleys are the rubbish-heaps of 

 small valley or corrie glaciers. At times the amphitheatre form is double, one being 

 above the other, indicating in all probability a halting-stage during the recession of 

 the ice. Similar landscape forms frequently occur in the valleys heading the sounds 

 of western Otago. 



There appears, therefore, to be satisfactory evidence that the islands under- 

 went a glaciation of moderate intensity ; but there is evidence that at an earlier date 

 it was much more intense. No large terminal moraines are visible, but they are 

 probably submerged at the present time. Although the floor of Carnley Harbour 

 does not show from its soundings the form of a fiord, yet the sides in places — and 

 particularly the two large peninsulas, Circular Head and Musgrave Peninsula — 

 show the rounded shape which would be produced by the rasping action of a glacier. 

 No grooves or striae were observed ; but these could not be expected to be visible, 

 owing to the complete covering of peat. 



The harbours on the eastern coast north of Carnley give important additional 

 evidence of this glaciation. Norman's Inlet, one of the most beautiful of these, 

 reproduces on a small scale almost every detail of the great fiords on the south- 

 west coast of New Zealand (fig. 6). There are no overlapping spurs ; the valley- 

 sides run down steeply into deep water ; the upper end of the valley termi- 

 nates in a great wall-sided amphitheatre ; a typical hanging valley occurs on the 

 southern side of the inlet, pouring its water into the sea in a series of cascades. The 

 dividing-wall between the main valley and its tributary hanging valley ends in a 

 sheer bluff like that at Milford Sound, but on a smaller scale. Soundings show 

 that the sound is deeper near the head and shallower towards the entrance. It 

 therefore possesses all the characters of a fiord. Even more striking is Cascade 

 Inlet, just south of it. This is a double valley — one a well-marked hanging valley, 

 with a waterfall from which ships can fill their tanks directly from the falling water. 

 Its sides are precipitous, and the soundings show some remarkable features. The 

 upper portion of the harbour shows a submerged basin, with the soundings reaching 

 32 fathoms ; but it is crossed by a bar half-way down, where the soundings are only 



