SECTIONAL TRANSACTIONS.— E. 447 



buried deeply under the ice, showing glacial smoothing as a result of abrasive 

 action only. When some of the deeper Welsh and Scottish cirques are 

 examined they are found to reveal plucking and shattering on their head- 

 walls right down to the junction with the cirque floor, even when this face 

 exceeds 1,000 ft. in height. The plucking action that Johnson so vividly 

 described at the foot of the bergschrund appears to have functioned over 

 the whole of the head-wall. 



The evidence of Icelandic cirques suggests that an elaboration of the 

 bergschrund hypothesis might better account for the facts than the hypo- 

 thesis as first put forward by Johnson. During the height of summer 

 melting in Iceland the head-walls of the cirques frequently had a large 

 amount of melt-water pouring down them. This melt-water resulted from 

 direct precipitation, from the melting of winter snows clinging to the upper 

 portions of the head-wall, and in certain cases the supply was considerably 

 augmented by melt-water from a small ice-cap resting on the summit above. 

 The volume of water varied greatly from day to day. On reaching the n^vd 

 the water almost invariably melted its way under the ice, thus following 

 the rock face below the glacier. This melt-water was frequently noted 

 pouring down the head-wall where this was visible along the bergschrund, 

 when it was usually augmented by further melting in the ndvd above. 



In a cave which followed the rock face under an ice-fall of considerable 

 thickness it was found that even on a warm day the moisture on the rock 

 face was frozen hard. This and further evidence of a similar nature led 

 to the suggestion that much of the melt-water seen pouring down the head- 

 wall, when it gets a certain distance under the glacier, freezes, and that this 

 is particularly liable to happen at night and during cold spells. 



The foregoing evidence suggests the following sequence of events in 

 cirque erosion. Large areas of the head -wall are covered with moisture 

 when melting is active and much of this moisture subsequently freezes. 

 The resulting frost action rots the head-wall and loosens large and small 

 rock fragments. These are then removed by becoming incorporated in 

 the glacier as it moves slowly away from the head-wall. This action is 

 naturally most effective when the walls are steep, so that steep slopes, once 

 formed, tend to be self-propagating. The cirque therefore becomes 

 enlarged by the head-walls slowly eating back into the mountain mass. 



Symposium on English ports and estuaries in their geographical setting 

 (10.45). 



Mr. F. H. W. Green. — Southampton — hydrographic factors. 



In discussing the economic geography of a seaport such as Southampton, 

 it is very necessary that the facts of its physical hydrography be accurately 

 and systematically stated. This is comparable to the accurate description 

 of the geomorphology of a region on land, which is recognised as a necessary 

 prelude to any discussion concerning its economic development. 



The tide is one of the more important aspects of port hydrography, and 

 in the case of Southampton it is of peculiar interest and significance. The 

 unusual complications, which make themselves manifest here, necessitate 

 more careful study than is demanded in assessing the significance of the tide 

 at most British ports. The existence of the complications is, on the other 

 hand, of great practical value in that it enables methods to be devised of 

 forecasting deviations from ' predicted ' tidal times and heights ; some of 

 these have already proved themselves to be surprisingly accurate. An 

 analysis of the tides at Southampton also suggests more accurate methods, 



