684 SUBANTARCTIC ISLANDS OF NEW ZEALAND. [Geology. 



east Harbour. A few feet only separate the top of the western chff from the easterly 

 slope. It is only in North-west Bay that any appreciable proportion of the rainfall 

 escapes to the west coast. Two small streamlets enter the bay, and deliver the 

 drainage from the west and part of the north slopes of Mount Dumas and from the 

 Menhir. The water from the east of Mount Dumas and from the north of Mount 

 Eboule, with that from the south of Filhol Peaks, enters Monument Harbour ; while 

 water from the east of Mount Honey, as well as from the south of it, falls into South- 

 east Harbour. The largest stream seen was one which falls into the head of North- 

 east Harbour. Its channel is 10 ft. deep, and it has the same width, with sides 

 formed of peat, and with a rocky floor. Except on the steepest slopes, the smaller 

 streams flow over peat floors. The peat is constantly slipping slowly from the hill- 

 slopes on all sides towards the bottom of the valleys. The main valleys have straight 

 or gently curved courses, with evenly contoured sides and semicircular heads. A 

 slight exception is found on the northern side of the valley which leads from the head 

 of the North-east Harbour, for here, below the terraces of lava, are to be seen larger 

 but less-marked terraces, sloping towards the head of the harbour at an angle of 

 about 5 degrees. There is one exception to the general form of the valleys as out- 

 lined above. The stream which flows from the eastern side of Mount Honey has 

 a head of the usual amphitheatre shape, but its floor is more encumbered with large 

 angular boulders than that of others, perhaps due to the fact that peat does not 

 grow so readily at the altitude of 800 ft. This is probably the valley referred to by 

 Hector : " At an altitude of 800 ft. clear evidence of the former existence of a true 

 corrie and moraine of the first order was obtained."* Its floor drops suddenly, 

 and the stream flowing in it falls over bare rock, into which a small groove only 

 has been eroded. There is very little detrital matter where streams enter harbours. 

 The soundings of Perseverance Harbour show that the form of the valleys which 

 enter the head of the harbour is maintained within it. The floor is nearly fiat, and 

 the depth gradually increases from the heads of the coves to beyond Terror Point, 

 where it is about 25 fathoms, though as much as 27 fathoms is recorded at one spot. 

 At the entrance the depth is 20 fathoms. This shows an overdeepening in the middle 

 part of the harbour. The soundings in North-east Harbour increase regularly to 

 the entrance. In all cases these soundings are taken from those recorded on the 

 Admiralty charts. Those given in Filhol' s plan differ but slightly from them. 



A noticeable feature in all the natural sections is the absence of soil between 

 the rock and the peat. No material was noticed which could be regarded as a 

 product of ordinary processes of weathering by atmospheric action. 



It is believed that all these features of surface form and structure can be most 

 readily explained as the effect of severe climatic conditions, which gave rise to re- 

 latively large glaciers. If it is sought to explain the form of the valleys by appeal 

 to stream erosion a prolonged action must be supposed, for the wide valleys are 

 relatively flat-bottomed. Prolonged action would have changed the juvenile features 

 of the valley east of Mount Honey. If the valleys were formed by stream erosion 

 the climate must have been somewhat different, for at the present time it favours 

 the growth of such vegetation as to produce peat, which protects the rock from 

 ordinary waste and decay. However, the growth of peat has been continuous ever 



* Hector, Trans. N.Z. Inst., vol. xxviii, p. 736. 



