446 JOURNAL OF SCIENCE. 
last swell overlooking the Tasman river. We had now to descend 
about 200 feet, and again came upon the track leading up the 
river bed. This river bed of the Tasman, over two miles wide, 
is a broad sheet of coarse gravel, through which the river mean- 
ders in countless channels, between which are often dangerous 
quicksands. We drove along over marshy flats, on which numer- 
ous seagulls had their nests (one of the young seagulls we 
afterwards met high up on the glacier, winging its flight over the 
snowy range to the west coast), then across river channels, 
and then over wide tracts of gravel. Right before us, rising 
abruptly from the river bed, in the point where the valley forked, 
was the great mass of Mount Cook, its icy peak glittering like a 
pinnacle of frosted silver against the deep blue sky. On either 
side the mountains rose from the flat valley with the same 
abruptness, and the terminal face of the Hooker and Tasman 
glaciers closed in the end of the two branches into which the 
valley divided to the right and left of Mount Cook. This flat 
river bed, with the mountains rising from it abruptly, and from 
margins as sharply defincd as the shores of a lake, is so typical 
of all the mountain valleys we saw, that we may ask, What is 
the cause of a feature so distinctive ? I believe the low level to 
which the glaciers descend, and the consequent short incline of 
the rivers, is a sufficient cause. The terminal face of the Tasman 
glacier is, according to Dr Haast, only 2456 feet above the sea ; 
while the mean of four observations, taken in as many days by 
myself, makes it 100 feet lower; and its river descends to the 
sea level by a fairly uniform incline of about 25 feet to the mile. 
If the river had a greater depth to descend before reaching the 
level country or sea level, it would erode a deep ravine-shaped 
bed, like those so common in the European Alps. High up on 
the mountain slopes, on the side of the valley opposite to where 
we travelled, were the most remarkable series of terrace forma- 
tions I ever saw, their level being quite 500 or 600 feet above the 
present river, and their edges sharply defined. Dr. Haast con- 
siders that they form part of the margin of an ancient lake, 
which was dammed up by a glacier crossing the valley lower 
down during the last great glacier period. 
Accepting, in part, this interpretation of the phenomena, 
several interesting questions follow, which we will try to answer : 
What river or rivers fed this lake? Was it the Tasman? The 
present source of the Tasman being about 200 feet lower than 
the terraces, would be below the level of the ancient lake, so that 
it could not have been the feeder, unless the lake existed in an 
inter-glacier period, when the climate was milder, the ice-cap 
smaller than at present, and the source of the Tasman higher up 
the valley. Supposing it was not filled by the Tasman river, it 
seems to follow that, at the t’me of the existence of the lake, the 
great trunk glacier formed by the junction of the Hooker and — 
Tasman glaciers must have filled up the centre of the valley, and 
extending far away down beyond the terraces, formed the dam 
which banked up the drainage of the hills above the terraces, 
