( '64 ) 



fince thofe parts, as far as the eye can per- 

 ceive, fliew no figns of the adion of fire. 

 Fourthly, That thefe three Ifiands, and per- 

 haps hkevvife Stromboli, are fituated in the 

 vicinity of analogous but primitive rocks* 

 The perfedl refemblance of the fhoerls and 

 feltfpars in thefe rocks, both in thofe that 

 have fuflered change from the fire and thofe 

 that have not, is a demonftration that thefe 

 cryftallizations have not been taken up by 

 the lavas when they flowed in currents, 

 nor formed in them at the time of their 

 congelation. 



From the very beginning of this work, I 

 have adduced fads of the fame nature, 

 which I have the fatisfa£lion to confirm by 

 the prefent ; and which become of the 

 greater importance, as I have lately read 

 that a naturaUft of eminence inclines to be- 

 lieve that the fhoerls of lavas are formed 

 when the latter ccndenfe and become cold, 

 becaufe then the homogeneous molecules 

 feparate from the heterogeneous, and unite 

 by affinity in fmall cryftallized mafles. 



This 



