BRECKNOCKSHIRE AND EAST GLAMORGANSHIRE. 53 



smallness of the fall between the levels of the resting-places of these 

 erratics and that of their parent rocks, (2) at Cwm Sarn by the 

 want of conformity between the lie of the longest axes of the 

 erratics and the trend of the grooves on the neighbouring rock- 

 surface. This point, however, needs to be more fully worked out 

 before it be accepted as true. (3) The contact in this group of 

 perfectly angular and rounded boulders can be most easily explained 

 on the hypothesis of an ice-foot, transporting a mixed load of water- 

 worn or glaciated boulders from shore-lines and angular rock-frag- 

 ments from overhanging cliffs. jSTo definite conclusion can as yet 

 be arrived at with reference to the rest of the boulders, with the 

 exception perhaps of those composed of Old Red Sandstone, in South 

 Brecknockshire, the frequency with which these erratics occur 

 angular, and the certainty of land-ice having once radiated from 

 the Beacons, justifying the inference that many of them are the 

 relics of old moraines. 



II. The Boulder-clay of South Brecknockshire is chiefly the pro- 

 duct of land-ice ; for (1) it is unstratified ; (2) its included stones 

 are intensely glaciated ; (3) angular stones greatly predominate over 

 rounded ; (4) in one case the largest of the included stones were 

 observed to lie with their lengths parallel to the lines of glaciation 

 on the underlying rock-surface. 



III. The striated rock-surfaces of South Brecknockshire have 

 been formed by land-ice, descending existing valleys, from the 

 Beacons, at least as far south as the edge of the Coal-measures. 

 South of this line the striated rock-surfaces present a more complicated 

 problem. Those at the Craig-y-gaer and at the top of the Rhondda- 

 fawr valley can hardly be attributed to the glaciers of the Beacons ; 

 and yet the former were certainly, and the latter probably, caused 

 by land-ice. 



In the coal-basin the striae having a S.E. trend and occurring at 

 high levels may be due to the glaciation of an ice-sheet coming 

 from the N.W., or the grounding of icebergs; the mamillated sur- 

 faces, however, point to land-ice as the agent. The striae which are 

 found at lower levels in the Taff and Bhondda valleys may be due 

 to (1) this N.W. ice-sheet, (2) the glaciers of the Beacons, (3) 

 glaciers having their origin in the coal-basin. 



The difference of direction in which the low-level striae run as 

 compared with the high-level, e. g. near Pont-y-pridd, may be due 

 to the land-ice conforming more and more to the trend of the valleys 

 as it became localized, or to their belonging to a glaciation altogether 

 more recent. 



The author has received much valuable advice from Professor 

 Prestwich. He has also to acknowledge the important help given 

 him by Dr. C. T. Vachell, M.D., Mr. W. Adams, F.G.S., and many 

 other members of the Cardiff Naturalists' Society. 



