390 Geology of 



the valley, small outlets Lave been in existence. They are, however, very 

 faint, and the lateral moraine has not been removed. It appears, 

 therefore, that the water melting during the advance of the ice, has 

 been leaking through the incoherent shingle bed f ormiug the floor of 

 the glacier, otherwise it would have certainly cut through the lateral 

 moraine, consisting all along of silt only, with rarely any boulder or 

 erratic blocks amongst it. The glacier channel is always stony, the 

 stones being either rounded or sub-angular, but in several instances 

 quite angular. Near the termination of the glacier channel none of 

 the blocks have lost their sharp angles. In a word, the whole 

 forms one of the most interesting geological phenomena I ever met 

 with — a glacier channel nearly three miles long and only about two 

 hundred feet broad, passing straight over a plain, ascending four 

 terraces together about 80 feet high, without furrowing to any depth 

 the alluvial shingle beds of the broad valley, and being bounded 

 at the same time by banks of silt. It is evident that a very large 

 quantity of ice must have existed higher up the valley from which this 

 narrow strip received the necessary vis a tergo to overcome all obstacles 

 in its way, the larger glacier after the separation of this branch 

 following without doubt the broad river bed where now the gorge of 

 the Eakaia is situated. (See Xo. 9, ground plan of Railroad glacier 

 on Plate Ko. 7.) 



The Eas-gitata G-lacieb. 



This glacier was also of considerable size, having been about 4S 

 miles long. It reached during its greatest extension several miles into 

 the Canterbury plains, crossing the front ranges — before the present 

 lower gorge was cut — by a saddle to the south of it. After its retreat 

 to higher regions a lake was formed behind the front range. This lake 

 in course of time was partly filled up by alluvium, partly drained by 

 the formation of the lower gorge. Morainic accumulations, numerous 

 glacier shelves, terraces and roclies moutonnees in the middle Eangitata, 

 bear ample testimony to the wide-spread glaciation of the country 

 during that remarkable period. The Eangitata glacier was anastomos- 

 ing with the united Eakaia-Askburton glacier by the large opening in 

 which now Lakes Tripp and Acland are situated, receiving after- 

 wards again a branch of the latter by Trinity Valley. After the 

 partial retreat, it continued to send a branch into the former men- 

 tioned broad opening, where several frontal moraines round Lake- 



