7Q PROCEEDINGS OF THE GEOLOGICAL SOCIETY. 



tend to form narrow and tortuous water-channels, such as are occa- 

 sionally met with in that deposit, and through which the water, 

 from the rapidity of its escape, would not act so fully as when filter- 

 ing slowly through the solid beds. It is easier also to drill a straight 

 hole in a mass which is uniformly compact and solid, than in one in 

 which weak points and cavities interfere with the regularity of the 

 drilling action, and I believe that it will be found that these sand- or 

 gravel-pipes are not usually connected with fissures, but that they 

 have generally worked their way into the solid chalk*. 



To understand this mode of operation, we have to suppose the 

 superincumbent strata constantly filled with water to a given level, 

 as happens with all water-bearing strata under certain conditions. 

 If these strata reposed upon an impermeable deposit, and the water 

 had no underground means of escape, the additions incessantly made 

 to it by the fall of rain would cause it to overflow the edges of this 

 natural reservoir and so carry off the surplus supply in the form of 

 springs ; but if the water had the means of escaping by any under- 

 ground vent, the ordinary hydrostatic pressure would cause it to take 

 that course in preference to the other. 



As, therefore, in the case before us, sands or gravel saturated with 

 water repose upon a deposit, which, notwithstanding its not being 

 freely permeable or allowing of the escape of the water in bulk, but 

 rather holding it up, still imbibes water readily, and transmits it, 

 although with extreme slowness, to the lower level which it seeks ; it 

 would follow that the surface of the chalk being exposed to the action 

 of the water, with which these overlying beds are constantly saturated, 

 would imbibe and transmit it downwards, at first through its mass 

 generally, but ultimately passing, probably, into some of the nu- 

 merous fissures which traverse this deposit, and which would afford 

 it a more rapid passage. But as rain-water contains carbonic acid, 

 which carbonic acid would not be lost by the filtration of the 

 water through these quartzose sands or gravel, a certain proportion 

 of that part of the chalk through which the acidulated water first 

 passed must necessarily be dissolved and carried away in solution. 



Under these circumstances, any slight original depression on the 

 surface of the chalk tending to give a direction to the water ; any 

 point where the texture was looser than usual, and presenting there- 

 fore less resistance ; or any greater permeability at some given point 

 of the superincumbent bed allowing a freer access at that spot, would 

 tend to facilitate and direct the passage of the water at and to those 

 places, and the inevitable consequence would be to establish neces- 

 sarily a greater wear there than elsewhere. By these means, the chalk 

 at these points would become more porous and less resisting, and cer- 

 tain water-channels there would soon be fixed, which, as they offered 

 the readiest passage for the water, would draw it off from the adjacent 

 portions of the superincumbent strata. This action once established 



* Still it is quite possible that slight crevices and cracks may frequently have 

 been predisposing causes : what I wish to point out, is, that the pipes are not 

 usually connected with faults and large fissures in the chalk, that they are not 

 dependent upon any lines of disturbance. 



