72 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. 



Oxfordshire, Bucks, and Hertfordshire. The general features de- 

 scribed above, and those given more in detail by Sir Charles Lyell, 

 may apply to them all. Care, however, must be taken to discriminate 

 between pipes produced by the slow percolation of water, and those 

 formed in part or wholly by the wear and battering to which the 

 chalk, as well as other strata not always calcareous, has been sub- 

 jected by certain drift-action. Such excavations are, however, on the 

 whole, shallower, more irregular, and do not show the same internal 

 structure as the true pipes. In this case many of the apparent 

 pipes are mere sections of deep furrows, ploughing up the surface 

 of the substratum, and filled with drift of various ageSo In one 

 of my papers recently published *, a case of this description inci- 

 dentally occurs where a mass of subangular ochreous sandy gravel 

 reposes on a deeply furrowed surface of impermeable clay and 

 siliceous sands f , 



§ 3. General Phcenomena. 



Such being the ordinary features exhibited by the sand- and gravel- 

 pipes, it now remains to consider when and under what general 

 conditions they were formed. As before mentioned, the percolation 

 of rain-water, holding in solution carbonic acid derived from the 

 atmosphere and the vegetable soil, from certain permeable strata 

 into the chalk has been the cause most frequently suggested. The 

 question, however, involves some further considerations which this 

 hypothesis does not embrace. These tubular excavations are evi- 

 dently to be attributed to some very general cause, and one operating 

 at different periods ; and it is with respect to the greater number a 

 cause no longer in action, for not the slightest subsidence or in- 

 dentation of the ground at the surface is perceptible above them, 

 although it is evident that whilst they were being formed the 

 superjacent strata were gradually depressed and indented (see figs. 1 

 and 2, pp. 69, 71). 



The Thanet Sands and the sands associated with the Reading and 

 Woolwich Series afford numerous examples of pipes passing down 

 into the underlying Chalk. These pipes are not only common in 

 the localities where the sands are in connexion with the main mass 

 of the Tertiaries and in all their detached outliers, but portions or 

 tail-pieces of such sand-pipes are also found on bare chalk-hills, above 

 which once extended those lower tertiary sands from whose mass 

 these pipes must necessarily have originally projected. This is suffi- 

 ciently evident in all the Chalk district of Kent. In illustration of 

 this point I have selected for a section the line of country between 

 Chatham and Crayford, along which instances of these unattached 

 pipes are in places numerous. The dotted lines mark the probable 

 extension of the Tertiaries before the denudation of the present val- 

 leys (see Section No. 2, PI. VI.). 



* Quart. Journ. Geol. Soc. vol. x. p. 88. 



t Considerable complication is often produced by drift-action on the surface 

 over the pipes, as the drift, being often local, can hardly in some cases be distin- 

 guished from the materials in the pipes. The inclination at which the section is 

 taken also gives rise at times to deceptive appearances. 



