166 GLACIOLOGY 



Pony Lake. This small lake, close to which our winter quarters were estab- 

 lished at Cape Eoyds, measured about 9 chains by 3 chains. Its surface is about 

 1 5 feet above sea-level. The water was somewhat .saline, partly through .salt spray 

 from the blizzards, partly through mineral matter dis.solved out of the surrounding 

 kenyte lava. It occupied a shallow rock basin excavated by the ice of the Great 

 Ice Barrier. The ice of this lake showed about ll inches of ablation between 

 April 18, 1908, and June 12 of the same year. The whole of the ice of this lake, 

 which is quite shallow, melted during the summer. At the time of our arrival early 

 in February 1908 the lake surface was recently frozen over. 



The general appearance of Pony Lake is shown on Plate XLIII. Fig. 1. 



LAKES AT CAPE BARNE 



As the Cape Barne lake district was 2 miles or more away from our Hut at 

 Cape Royds, detailed work was impossible during the winter. During the spring and 

 summer also sledging preparations and sledging prevented our spending much time 

 in scientific work near the camp, and consequently the lakes were not examined 

 nearly as carefully as those at Cape Royds, During our few visits to Cape Barne 

 the lakes were roughly surveyed for adding to the Plane Table Map. Most of the 

 notes on the Cape Royds lakes would apply to those at Cape Barne, at any rate 

 as far as superficial characteristics were concerned. The largest of these lakes 

 were situated in roughly parallel grooves running N.N.W. to S.S.E., and were 

 named by us Terrace Lake, Deep Lake, and Sunk Lake ; the smaller circular 

 depressions, when ice-filled at all, were occupied by small tarns. 



It was in one of these small tarns that our best example of the formation of a 

 convex surface was seen. The method of formation seemed to be here as follows : — 



The ice expanding as it was formed was forced upwards in the centre, and thus 

 a convex surface was produced ; this was exaggerated by the formation of each 

 successive layer, until at last the pressure on the top layers was sufficient to over- 

 come the cohesion between the molecules, and a crack resulted. As the process 

 continued this crack would become wider and wider, and would affect a greater 

 thickness of ice, until, when finally the water of the lake was all frozen, the lake 

 would be occupied by a bi-convex lens of ice with cracks, wide at the top and 

 tapering away to nothing below. In some cases this process was modified by the 

 cracks traversing the whole depth of the ice and tapping the water, when recement- 

 inof of the cracks took place. This convex surface would sometimes be partially 

 or completely removed by ablation during the winter, and a level ablation-rippled 

 surface substituted. 



Terrace Lake. Terrace Lake occupies the longest valley in the Cape Barne 

 district. It is about 700 yards long, and is almost divided into three lakes by two 

 hard ridges of kenyte. Its surface is about 180 feet above sea-level. It is sur- 



