286 LETTEKS ON NATURAL MAGIC. 



has been very successfully applied by Sir John Herschol 

 to explain the subterranean sounds with which every 

 traveller is familiar who has visited the Solfaterra near 

 Naples. When the ground at a particular place is struck 

 violently by throwing a large stone against it, a peculiar 

 hollow sound is distinctly heard. This sound has been 

 ascribed by some geologists to the existence of a great 

 vault communicating with the ancient seat of the volcano, 

 by other writers to a reverberation from the surrounding 

 hills with which it is nearly concentric, and by others to 

 the porosity of the ground. Dr. Daubeny, who says that 

 the hollow sound is heard when any part of the Solfaterra 

 is struck, accounts for it by supposing that the hill is not 

 made up of one entire rock, but of a number of detached 

 blocks, which, hanging as it were by each other, form a 

 sort of vault over the abyss within which the volcanic 

 operations are going on.* Mr. Forbes, who has given 

 the latest and most interesting description of this singular 

 volcano,f agrees in opinion with Dr. Daubeny, while 

 Mr. Scrope J and Sir John Herschel concur in opinion 

 that no such cavities exist. " It seems most probable," 

 says the latter, " that the hollow reverberation is nothing 

 more than an assemblage of partial echoes arising from 

 the reflection of successive portions of the original sound 

 in its progress through the soil at the innumerable half- 

 coherent surfaces composing it : were the whole soil a 

 mass of sand, these reflections would be so strong and 

 frequent as to destroy the whole impulse, in too short an 

 interval to allow of a distinguishable after-sound. It is a 

 case analogous to that of a strong light, thrown into a 

 milky medium or smoky atmosphere ; the whole medium 



* Description of Volcanoes, p. 170. 



T Edinburgh Journal of Science, N. Series, No. i. p. 124. 



* Considerations on Volcanoes, and Edinburgh Journal of Science, 

 No. xx. p. 261, and No. xiv. p. 265, 



