Geological Survey of Canterbury. 131 



tation, with its various tinted foliage, clothes the mountain side. 

 Eor two miles more we kept along the eastern side of the valley, 

 travelling mostly on the slopes of huge shingle-fans, and camped under 

 the shelter of a dense group of Phyllocladus alpinus and Dracophyllum 

 longifolium, the north-west winds blowing down the valley with great 

 fury. 



Xext morning we started with the dawn, hoping to reach, if possible, 

 the head of the valley. Pour miles from our camp the valley assumes 

 all the characteristics of a gorge, in which the river leaps incessantly 

 over huge blocks. After another mile of perpetual climbing over such 

 boulders and along taluses of debris, the valley received an important 

 addition from the west, containing nearly as much water as the main 

 river above. This torrent, which we had to cross near its junction, 

 flows in a deep narrow gorge, having the appearance of a deep cleft, 

 which has rent the chain from top to bottom. The higher we ascended, 

 the more the valley narrowed and assumed a rugged appearance ; at 

 the same time, the vegetation became strictly alpine, and many of the 

 plants were still in full bloom, filling the air with a delicious fragrance. 

 A mile below the glacier a large avalanche lay across the river-bed, 

 forming a snow bridge from side to side, through which the water had 

 formed a tunnel. Two very prominent peaks rose conspicuously above 

 us, of which the south-westerly one, Mount Tancred, sends a glacier 

 down to the valley ; its terminal face I ascertained to be 3788 feet above 

 the sea. Another majestic peak, Mount Carus, lies in a north-easterly 

 direction behind the former, but owing to the great steepness of its 

 sides the snowfields on it are of much smaller dimensions. Another 

 glacier descends from the ridge connecting the two peaks in a deep 

 gorge, and terminates a quarter of a mile above the glacier previously 

 described. Returning on the 24th of March to the junction of the 

 Mathias, I devoted two days to the examination of the slopes of the 

 Mount Rolleston and Mount Algidas ranges, and in preserving and 

 putting in order my collections. I arrived on the 27th March at Groat 

 Hill accommodation house, where I had my horses shod, and deposited 

 the collections, which had already augmented so considerably that 

 they formed quite a load for one horse. 



"We started on Wednesday, the 28th of March, up the Wilberforce, 

 and kept along the western base of Groat Hill, on the large alluvial 

 deposits brought down by the Boulder stream and Kakapo creek. 

 Before reaching the junction of the Kakapo I observed the 

 remains of a large moraine, crossing the valley and cropping out 



