452 



PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [May 28, 



the gorge at a, which used to flow regu- 

 larly clown the fan from a to d, may, 

 indeed, for a time flow nearer the moun- 

 tains down another of the radial water- 

 courses to a part of the fan- edge where 

 the river has not eaten its way in; but 

 it is sure some time or other to find its 

 way over a part of the fan that leads to 

 the cliff ; Yover this, then, it will fall ; and 

 it will soon cut a channel for itself in the 

 cliff-edge, and continue to cut back that 

 channel so as to modify the shape of its 

 bed as far back as a or even further, into 

 the ravine itself. This action however, 

 may, be interrupted by some cause origin- 

 ating at or near the mouth of the ravine, 

 which shall deflect the river-course and 

 lead it either, as before, down an unbroken 

 part of the fan or to another portion of 

 the cliff, in which last case a similar action 

 would be set up there. Thus we often see 

 gullies cut in the cliff-edge where the 

 stream had for a time flowed, but from 

 which it had been stopped above before 

 any very deep channel was made. Ulti- 

 mately, however, the stream is sure to 

 deepen one course for itself so much and 

 so far back that all further wanderings 

 are prevented, and for the future it will 

 flow out along a deep gully between cliffs 

 of the fan-alluvium. 



Yet another effect is produced : over 

 this newly cut gully, while it is forming, 

 will come with the water the usual de- 

 tritus ; and that will (with the addition of 

 the waste of the fan itself involved in the 

 cutting of the gully) fall and settle at the 

 foot of the cliff in a more or less modified 

 talus ; as this rises, and the gully deepens, 

 the talus and the new stream-bed will 

 meet, and the stream will then flow out 

 over the talus, and eventually, by still 

 adding material, lessen the angle of the 

 talus-slope and make a new fan with 

 nearly the same inclination as the old one 

 had, projecting in front of the fan-cliff. 

 In the diagrammatic section, fig. 8, the 

 lower unbroken line from a to b denotes 

 the position of the new stream-bed in 

 the gully, while the continuation of it 



