Volcanos and Earthquakes. 363 



hot springs, by evolving steam from their interior, and why 

 they never appear to give issue to that class of thermal waters 

 which I have noticed in Ischia, as being unaccompanied with 

 gaseous products." 



A spring arising from beneath leads us to conclude that me- 

 teoric water penetrates through clefts which communicate low 

 down with the former. The experience gained in boring arte- 

 sian wells shews that a succession of strata is most favourable 

 for such processes, and from causes easily explained. In what 

 are called primary rocks, however, no such alteration of strata 

 is found, because they are not stratified. The usual occur- 

 rence, viz. the flowing of meteoric water down inclined surfaces 

 of stratification which appear at elevated situations, and the 

 rising of this water, by means of natural or artificial channels, 

 after having been forced down to a more or less considerable 

 depth, cannot then happen in unstratified rocks. It appears, 

 nevertheless, that there are granitic rocks traversed by clefts 

 more or less perpendicular, and communicating low down. 

 Thus at Aberdeen, in Scotland, water has been drawn by boring 

 in granite 180 feet below the surface, which, according to Ro- 

 bison, came from a cleft filled with sand and gravel, and rises 

 six feet above the level of the earth.* Such a communication 

 of the clefts low down, must, however, occur but rarely. 



If the primary mountain rises above its environs and the 

 clefts at its base lie exposed, then will the springs flow out of 

 the clefts. Such an origin of springs, which are not naturally 

 rising springs, is often observed at the foot of basaltic and tra- 

 chytic cones, &c. 



On the other hand, on the limits between stratified and un- 

 stratified rocks, where the latter have traversed the former, 

 and where channels extending to a great depth have been form- 

 ed in consequence of the contraction of the traversed masses 

 during their cooling, circumstances favourable to these rising 

 springs exist, and it is easy to conceive, therefore, that thermal 

 springs may be found on the limits of these interrupted masses, 

 but not in their interior. 



Let us imagine a stratified chain of mountains consisting of 



* Compt. Rend. 183 No. 24, p. 675, and t ii. No. 20, p. 683. 



