Mr Hugh Miller on River- Terracing. 281 



with the decreasing force of the water.* The gravel-bank, 

 in fact, is growing by a process of accretion ; coarser mate- 

 rials (with finer entangled among them) are added to its 

 lower parts nearest the thread of the current ; finer materials 

 fall from the slackened grasp of the water further away from 

 it ; and the first tufts of grass, and first lodgment of sand, near 

 the edge of the alluvial meadow above, may appear together, 

 — the sand helping the grass to grow, and the grass entang- 

 ling more sand. An instance may be cited from the Vale 

 of the Coquet (Northumberland). While an alluvial curve of 

 that river has been slowly pushing through 80 yards in 18 

 years (as determined by the position of some boundary-stones, 

 and the age of the Ordnance map), a growth of gravel has 

 kept pace with it from the further side ; what was water be- 

 fore is gravel now, and what before was gravel is a flat of 

 scanty pasture, receiving earthy top-dressings from floods, and 

 harbouring, doubtless, an increasing colony of soil-forming 

 worms. 



According to the method just described, the formation of 

 at least the gravelly parts of river alluvia is not a horizontal 

 deposition of layers, but an accretion upon a slope, the finer 

 gravel covering the coarser by a kind of overlap. The de- 

 posits of sand and loam that generally (sand first, loam above) 

 succeed, lie in layers that fringe out and overlap, according 

 to the height of the floods that formed them. 



All this-f- is not without its bearing upon inductive 



* This assortment of materials stranded on the way down stream will be all 

 the better understood from a glance at the following table, showing the trans- 

 porting power of water in motion at different rates : — 



water will just begin to work on fine clay. 

 ,, will lift fine sand. 

 , , will lift sand as coarse as linseed. 

 ,, will sweep along fine gravel. 

 ,, will roll along rounded pebbles 1 inch 

 in diameter. 

 ,, 36 = 2-045 ,, will sweep along slijipery angular 



stones of the size of an egg. 

 D. Stevenson, Canal and River Engineering, p. 315. 

 f Although so simple as almost to require an apology for treating of them, 

 the processes above described are not universally understood. The author of 



