LAKES AT CAPE ROYDS 155 



of the lake shores was quite gentle. This steep rock slope is seen to the extreme 

 left of the lake shore in the photograph, Plate XLII. Fig. 2. It was near this 

 knoll, and with the object of tapping the deepest part of the lake, that our first 

 trench was made. 



Another important feature about Clear Lake was the terraced condition of the 

 slopes immediately above the present lake level. At the north-west end especially 

 there were well-marked traces of terraces in the moraiuic gravel ; these terraces 

 were each from 1 to 2 feet high, and may mark periods of rest during the recent 

 recession of the lake. 



First Trench, cut on March SOth. This trench was, as mentioned above, near 

 the kenyte knoll, and there proved to be then only 4 feet 3 inches of ice. The top 

 4 inches of ice is coarsely crystalline, with wide opaque divisions between the 

 more or less hexagonal prisms, and may be the result of the conversion into ice of 

 the lower portions of snow-drifts by the thaw, such conversion having taken place 

 in crystalline continuity with the prismatic ice formed each year by the re-freezing 

 of the water of the lake. The individual crystals are not at all regular in shape, 

 being often most intimately interlocked, and frequently arranged in curious patterns, 

 with a peculiar radial structure, as if they had grown outward from a central prism 

 or group of prisms. 



This roughly radial arrangement of prismatic ice seems quite common in the 

 older moraines on the Ferrar Glacier, and there is no doubt about its origin there, 

 for it is formed in almost every case by the freezing of the thaw-water after a 

 boulder has thawed its way down below the surface. The Western Party saw in 

 such cases the process in all stages, from the boulder that is just commencing to 

 sink in, to the rock moraine where not a boulder is to be seen, and their former 

 presence is only indicated by these radial patches of prismatic ice in coarse hexagons. 

 The structure seems here to be distinctly a secondary one, for although in a few 

 cases hexagonal plates of ice were observed on the surface of thaw-water lying in 

 hollows in the ice, it was far more usual for the first layer of ice at any rate to 

 form as small acicular needles radiating from every projection, and the ice sheet 

 newly formed presented to the eye an apparently homogeneous structure. 



It seems therefore that in this latter case at any rate some molecular rearrange- 

 ment takes place, with extrusion of air as minute air bubbles outlining the crystals. 

 It is suggested that this is a secondary structure superinduced upon ordinary thaw- 

 water ice. In the case of the prismatic ice on the lakes it was more difficult to 

 see any preliminary stages. . The surface ice was either ordinary thaw-water ice, 

 or the hexagonal prismatic ice was there in its finished condition. It seems probable 

 that in the lakes, where the freezing is undisturbed by the movement of the water, 

 the ice forms in vertical hexagonal prisms, as was undoubtedly the case in the 

 Narrows of Blue Lake, where the separate prisms could be traced right through the 

 ice. The thawing of the snow-drifts then proceeds in crystalline continuity with 



