THE PHYSICAL HISTORY OF THE FJORDS OP NEW ZEALAND. 155 



Lake McKerrow or Kakapo, though now a freshwater 

 lake 15 feet above sea level, and separated from Martin's 

 Ba,j by an alluvial flat three miles wide, is included in the 

 above table for the reason already mentioned. 



After passing the bar at the entrance, the floors of the 

 fjords run approximately level to near tlie head of the sound, 

 when they rise either into a clift' face or into a very small 

 alluvial flat, the former termination being perhaps more 

 general. 



Origin of the New Zealand Fjords. — On the origin of these 

 ■ sounds there is little room, 1 think, for speculation. They 

 obviously are submerged river-valleys, and, as will be seen^ 

 from a reference to the map, show all the characteristics of 

 valleys so formed. Their arms, originally tributary streams, 

 generally join the main sound at a natural angle, and the 

 sounds themselves, while preserving the general direction, 

 yet show minor deflections ; both features characteristic of 

 mountain valleys. Further, the change from the main 

 sound nito the arm is not marked by any appreciable change 

 in depth, and consequently in the level of the old valley 

 floor. While at a date subsequent to their erosion and 

 anterior to their submersion, they were undoubtedly occupied 

 by glaciers, few traces indeed of glaciation are still pre- 

 served. The islands in the sounds, though they show no 

 rugged contours, at the same time show no appearanceof 

 Stoss- and Lee-seiten, nor do the smaller rocks fm-nish 

 any positive evidence. Still, the fjord walls in a few places, 

 notably at the Narrows, Milford Sound, and in Deas 

 Cove, Thompson Sound, show evidence of polishing and 

 striation by glacier action.* With regard to the latter 

 place. Sir .lames Hector remarks {loc. cit.): " The rock here 

 is a granitic gneiss, the hard surface of which has faithfully 

 preserved the grooves and polished surfaces caused by ancient 

 glaciers." The terminal moraines in the Cleddau Valley and 

 elsewhere also furnish evidence of the former existence of 

 glaciers, as also do the granite erratics, 15 to 20 feet thick, 

 found at the mouth of Preservation Inlet among slates,t and 

 which must have been brought down the sound by ice. 

 Possibly the scarcity of glacial markings may be due to the 

 land, now above sea level, being covered in the Glacier 



* Hector : Geological Exploration of the West Coast, p. 453. 

 t Hutton and Ulrich : Geology of Otago, p. 68. 



