EXPERIMENTS on WHINS70NE and LAVA. 71 



their fucceflion is diftinc"tly feen, the fection of each flratum 

 prefenting to the view part of a horizontal circle ; the whole 

 confifts of alternate layers of thin ftreams of lava, and very thick 

 beds of loofe frothy rapilli, which laft being thrown into the 

 air in a foft flate, had fallen in fhowers on the fides of the 

 mountain. 



In various places the regularity of this arrangement is in- 

 terrupted by certain vertical lavas, from two feet to ten or 

 twelve in thicknefs, which crofs the flrata juft defcribed in an 

 irregular manner, and pafs upward, without diftinclion, through 

 the folid beds, and through the loofe Ones. It immediately 

 occurred to us *, that thefe lavas muft have flowed in nfTures 

 of the ancient mountain ; and we accounted for them by fup- 

 pofing, that a melted ftream, flowing along the external furface, 

 had met in its courfe with one of thofe crevices which are 

 formed in all great eruptions, and had flowed into it fo as to 

 return again into the heart of the mountain. This conjecture 

 very nearly agrees with thofe advanced by M. Dolomieu, and 

 by M. Breislack, who both mention thefe vertical lavas of 

 Somma f. 



I have fince been induced to confider this phenomenon, 

 which formerly feemed to prefent only an amufing variety in 

 the hiftory of volcanic eruptions, as of the utmoft confequence 

 in geology, by fupplying an intermediate link between the 

 external and the fubterraneous productions of heat. I now 

 think, that, though we judged rightly in believing thofe lavas 



to 



* I saw this place in company with Dr J. Home in 1785. 



f M. Dolomieu conceives thefe lavas to have flowed over the lips of the crater, 

 (IJIes Ponces, p. 100.J ; M. Breislack, that they had firft filled the open cavity 

 of the crater, and from thence had flowed into crevices formed in its fides, " che 

 " una lava avendo riempita la cavita del cratere fi foffe infinuata per quelle fendi- 

 " ture," (Topografia Fifica delict Campania, p. 115. J. This laft mentioned work,, 

 published in 1798, contains many interefling and accurate defcriptions. Should the. 

 circumftances of the times permit, the author will have it in his power to follow out,, 

 with every advantage, the hints I have fuggefted. 



