RIVER DEVELOPMENT. 65 



during part, at any rate, of that of the Middle Terrace, it seems 

 to have flowed to the north of the Purfleet and Grays Chalk 

 ridge, subsequently returning to the south side. The deposition 

 of the Crayford Brickearth and the absence of any break between 

 the Middle and Low Terraces may be connected in some way 

 with these changes, and the present valley below Purfleet and 

 Dartford is, compared with that above, somewhat gorge-like in 

 character. 



No important changes in the physiography since the time of 

 the Low Terrace cold period can be detected. The variations 

 in the character of the Coombe deposits show that the distribution 

 of outliers of gravel and of Tertiary strata was already much the 

 same as to-day, and the same inference may be drawn from the 

 occurrence of palaeolithic implements on the Chalk plateau, since 

 they are found most abundantly near the present outcrops of the 

 Bullhead Bed and the gravelly margins of the Clay -with -flints. 1 



The existence of numerous dry valleys in the Chalk area has 

 been taken to indicate a change in the climatic conditions. Many 

 of the valleys occupied by streams can be followed far above the 

 springs at which those streams normally originate ; in some we 

 find ' bournes,' that is, streams which flow only at intervals, 

 when the natural water-level rises temporarily, causing springs 

 to be given out at various points further up the valleys. 

 Descriptions and statistics of these bournes are given in the 

 memoirs on Water Supply. 2 The point where a dry valley 

 becomes a wet one varies, therefore, according to the season, and 

 if the annual rainfall was formerly greater than now, a greater 

 length of the Chalk valleys would have been subject to erosion 

 by streams ; but on a porous stratum, such as the Chalk, valleys 

 cannot be initiated in this way. Clement Reid has suggested 

 that certain Chalk valleys may have been formed by the action 

 of summer rains after cold winters on the frozen and consequently 

 impervious surface of the Chalk. 3 Other writers have accounted 

 for these coombes by a cause which is active at the present day, 

 the removal of chalk in solution by percolating waters. Rain 

 water, running off the clay which frequently caps the high 

 ground, enters the Chalk by means of swallow holes and tends 

 to follow the natural joints in the rock; some of the chalk is 

 removed along these lines and in this way a ' master -joint ' is 

 formed, carrying an underground stream like those known to 

 occur in the Chalk. The constant widening and deepening of 

 this line of weakness at length produces a deeply sunk valley, 

 in which a stream appears at the point where the base-line cuts 

 the water -table. 4 A study of the ground plan of the valleys, 



1 Dewey, H., ' Surf ace -changes since the Palaeolithic Period in Kent and 

 Surrey,' Proc. Prehist. Soc. E. Anglia, vol. ii, 1915, pp. 107-116. 



2 ''Water Supply of Kent' (Mem. OeoL Surv.), 1908, pp. 54-63; 'Water 

 Supply of Surrey' (Mem. Qeol. Surv.), 1912, pp. 61-71; 'Records of London 

 Wells ' (Mem. Geol. Surv.), 1913, pp. 25-29. 



3 Quart. Journ. Geol. Soc., vol. xliii, 1887, pp. 369-371. 



4 See, for instance, Young, G. W., Proc. Geol. Assoc., vol. xix, 1905, pp. 

 191-193 ; Spicer, Rev. E. C., Geog. Journ., vol. xxxii, 1908, pp. 288-291 ; Chandler, 

 R. H., Geol. Mag., 1909, pp. 538, 539. 



C 4 



