182 Transactions of the Royal Microscopical Society. 



One or two analyses of obsidian and leucite are appended, to- 

 gether with a few remarks thereon. 



Analysis of Obsidian from Lipari. 



Abich, 'Vulk. Ersch.,' 184], 62 and 84. 



Silica 74 



Alumina 12 



Sesquioxide of iron 2 



Protoxide of iron 



Lime 



Magnesia 



Potash 5 



Soda 3 



Water and loss 



Chlorine 



99-67 



• Although this analysis indicates that protoxide of iron is absent, 

 yet one of an Indian obsidian, by Damour, shows 10 "52 of the 

 protoxide, but an absence of sesquioxide of iron. 



When we consider that the protoxide of iron present in magne- 

 tite does not amount to more than about 30 per cent., and when 

 we also consider that the crystals in the spherulitic obsidian from 

 Lipari are partly altered or entirely altered into peroxide, either as 

 hematite or limonite, we need not be surprised that the above 

 analysis yields no protoxide of iron ; and even if the magnetite were 

 only slightly decomposed, its sparse dissemination in this rock 

 would render the presence of protoxide of iron merely discernible 



in traces. 



Analyses of Leucite from Vesuvius. 



Klaproth. 



Silica 53-75 



Alumina 24-62 



Lime — 



Soda — 



Potash 21-35 



Sesquioxide of iron — 



Water — 



Bischof. 



57-84 



22-85 

 0-2 

 604 



12-45 

 0-14 

 0-59 



Comparing these with the foregoing analysis of obsidian, it will 

 be "seen that the chief constituents of leucite and of obsidian are the 

 same, but that the mineral and the rock vary somewhat in their per- 

 centage composition. Surely this identity of qualitative composition 

 may to some extent account for the similarity which appears to 

 exist between the structures described in this paper. 



Before we finally quit this subject, let me once more direct 

 your attention to the section of spherulitic obsidian. To my mind, 

 we have here not merely a microscopic section, but a geological 

 section of considerable significance. Let us suppose that the 

 cortical zone of our spherulitic band is about 25 miles in thickness ; 



