54 GLACIOLOGY 



Barrier, but this in itself is not evidence that the Barrier is aground. If we 

 examine the evidence further on the north side of the Barrier as to v^hether or not 

 there is a tidal crack the evidence is unsatisfactory, for all the sea ice had gone out 

 at the time of our visit, and the Barrier ice was, as already stated, traversed by a 

 network of cracks and crevasses, and was intersected with what we called ice 

 barrancas and ice dongas. The origin of Relief Inlet (see Plate IX.), as weU as of 

 some of the ice barrancas and dongas, is almost certainly to be sought in shearing, 

 the result of the forward movement of the Drygalski Barrier. 



Relief Inlet has the form of a long, wide curving natural dock in the Barrier. 

 Its general appearance is shown in Plate XII. Fig. 1. 



The sea under the glacier near here is over 600 fathoms deep. The glacier is, 

 therefore, there, probably a piedmont afloat. 



In February 1909 there was open water at the bottom of the inlet for about 2 

 miles up from its mouth ; for the remaining 4 miles of the inlet, examined by us, it 

 was floored across with sea ice, very much cracked, with sea water showing through 

 the cracks and numerous seals lying on the ice. Its general appearance at the spot 

 where we made our unsuccessful attempt to cross it on the return journey from the 

 Magnetic Pole area is seen in Plate XII. Fig. 2. 



This ice barranca, a continuation shorewards of Relief Inlet, is about 150 

 yards wide and 50 feet deep. It will be noticed that it curves to the left to- 

 wards Cape Philippi, which is the northern boundary of the Drygalski Barrier 

 on the coast. 



At the point where we attempted to scale the cliff it was for its upper part 

 wholly formed of snow, that is, for fully 20 feet. Perfect curtains formed of icicles 

 concealed the middle part of the cliff" from view, and its base was hidden by drift 

 snow. It is difficult, therefore, to say whether this old snow was resting on old sea 

 ice or on elacier ice. Relief Inlet, and the ice barranca in which it terminates 

 inland, represent one of many shear planes which mark off" the Drygalski Piedmont 

 from the Nansen Piedmont. 



In travelling from a point a short distance to the north of this barranca, on our 

 outgoing journey to the Magnetic Pole Plateau, we passed a number of what we 

 called "ice dongas." These had the appearance of small trough faults and step 

 faults. Their appearance is shown in the following diagram (Fig. 12). 



For some time before reaching the first ice donga Mawson and Mackay reported 

 that sea water could be seen at the bottom of some of the cracks in the undulating 

 surface of ice and snow over which we were sledging. The dongas, like the 

 larger barrancas, are probably formed by shearing movements. Occasionally, but 

 rarely, we passed by an ice knob some 10 feet above the general level of the plain, 

 and they reported that they could see ice to a depth of 30 to 40 feet in the sides 

 of a crevasse running through one of these knobs. After crossing this ice donga 

 we traversed four pressure ridges, each of them with a small downthrow of 10 



