484 GEOLOGY OF THE SOUTH EAST OF DORSETSHIRE. 



tainous declivities of Studland Heath by a valley. Now such 

 is actually the case, for between the chalk and Studland there 

 is a deep diluvial excavation, which, in short, is continued 

 all along under the chalk, thereby insulating all Studland 

 Heath and Studland itself by dry straits, one of which now 

 affords a bed for a winter stream that finds its way through 

 one of the cracks in the cliffs into the sea. This fact would, 

 I know, be used differently by some geologists, who contend 

 that running water scoops out its own channels, even in the 

 hardest rocks, and, therefore, say they, sand but lightly ag- 

 glomerated must give way. Thus, the chines along the shore 

 of Poole Bay, are by Mr. Lyell said to be the result of the 

 streams that flow through them to the sea. If so, of course, 

 the Studland ' cracks ■ or chines have no right to be deemed 

 worth notice. But how stands the case ? A violent and pow- 

 erful torrent, bearing with it gravel and fragments of angular 

 rock, tosses these extraneous matters about in the hollows of 

 its bed, and they, acting like a mechanical machine, bear 

 away the moistened bed, till they cause that bed to descend 

 deeper and deeper in the solid rock below ; — and, therefore, 

 it is said, a sluggish stream must, of necessity, bore away in 

 sand with less trouble and more effect ! 



Now, I am not unaware that there may be cases found, 

 where the torrent has assisted in eating out a deeper channel 

 to a certain limited extent, — but I am not satisfied with the 

 assertion that this extent may be unlimited. The river Si- 

 oule in France is quoted as an example. That river now runs 

 at a level through nearly vertical walls of basalt and gneiss, 

 much below what it formerly did ; and this is shown by a 

 ledge of gravel much above its present bed. This gravel-de- 

 posit marks a period, it is assumed, when the river had only 

 eaten down so deep in the solid rock. It is said, that the 

 Sioule has cut through more than 100 feet of compact basalt, 

 and at least 50 feet of gneiss. 1 



But if the theory of these stone-eating waters be tiue, there 

 ought never to have been any gravel above, left on any ledge, 

 or else there ought to be a slope of gravel all the way down. 

 The river has descended certainly, but it must have suspend- 



1 See Messrs. Lyell and Murchison ' On the excavation of valleys, as il- 

 lustrated by the volcanic rocks of Central France ; ' G. P. i. 39, and Edin. 

 Phil. Journal : also Prof. Sedgwick's Address to the Geol. Society, Feb. 

 19, 1830, for facts and comments. After giving a luminous account of the 

 different modes of excavation, the latter distinguished author and observer 

 sums up with an allusion to the Auvergne rivers. These are great authori- 

 ties, and it maybe presumptuous to dispute their judgment, — but geologi- 

 cal doubts often lead to geological truth. 



