PRESTWICH — SAND- AND GRAVEL-PIPES. 



n 



would be maintained, and, in consequence, the original small water- 

 channels would be gradually worn larger by the continued solvent 

 action necessarily kept up by the continuous supply of water, re- 

 ceiving into their cavities, as they formed, the loose and yielding 

 materials of the superincumbent mass, and which, from the extreme 

 slowness of their fall and from all their parts being kept, by the uni- 

 formity and steadiness of the superincumbent pressure, in their original 

 relative positions, would, as it were, be merely stretched out, and 

 would conform almost exactly to the irregular surface of the pro- 

 duced cavities* (see fig. 1, p. 69). 



Further, the peculiar funnel-shaped and cylindrical forms of these 

 pipes admit of ready explanation on this hypothesis, for when any 

 point on the surface of the Chalk, owing to the causes above men- 

 tioned, gives a freer passage to the water than it can obtain through 

 the adjacent portions of the strata, then, the superincumbent mass of 

 sand or gravel being of uniform or nearly uniform texture, the water 

 would flow towards this channel from all points of the circumference, 

 and cause therefore a nearly equal wear on all sides, tending, as the 

 cavities deepened, to give them a circular form more or less perfect. 

 At the same time, as the hollow became deeper, the body of water 

 passing down the centre of the pipe would be greater in proportion to 

 the surface of the chalk to be acted upon than that which would fall 

 upon its sides, and a more rapid wear would take place on the lower 

 or central part of the tube than on the upper ; also the water passing 

 down the centre of the pipes would retain its erosive action undimi- 

 nished by previous contact with the chalk. 



The consequence of these conditions would necessarily be to trans- 

 form the original hollow first into a funnel-shaped cavity, and then, 

 as it got deeper, into one gradually assuming a more tubular and 

 cylindrical form. For if we divide a line, drawn through the centre 

 of the horizontal section of the y^^^^ 4^ 5^ ^ ^.-Diagrams illus- 



trative of the Structure of Sand- 



pipes. 



top of a pipe, into three equal 

 parts (Aa, ah, bB, fig. 4), and 

 carry down two perpendicular 

 lines from a and b, until they 

 meet the sides of the pipe at c 

 and d in the vertical sections, 

 figs. 5 & 6, it is evident that in 

 fig. 5 the relative dimensions of 

 Ac, cd, and dB are very nearly 

 the same, the line cd being very 

 little less than cA or dB ; still 

 the difference is suflicient, sup- 

 posing equal quantities of water 

 to pass in equal time through 

 the equal widths Aa, ah, bB, to 



* The comparative fluidity of the sands produced by their saturation with water 

 would necessarily facilitate this operation. The vibrations caused by earthquake 

 movements might also co-operate from time to time, especially where any tempo- 

 rary obstruction had caused a stoppage of the descending materials, in maintaining 

 a continuous fall into the increasing cavities. 



