Suppbsed Identity of Pumice with Lava, &c. 387 



The same chemist extracted from the Amazon stone. 



Silica 62-83 



Alumina IJ'OO 



Potassa 16-00 



Lime 0-30 



Oxide of iron Q-IO 



96-23 

 The Feldspath compacte ienace of Haiiy, or Sanssurite, con- 

 tain-'j according to Saussure, 



Silica 440 



Alumina 30-0 



Oxide of iron 12-5 



Lime 40 



Soda 6-0 



Manganese, potassa, ^ .^.j. 



V'' 



and loss 



100 



AndKlaproth found that the Klingenstein porphyr or sonorous 

 porphyry (var. P. compacte) was principally composed of silicas, 

 alumina, and soda. 



The quantity of alkali in pumice is much less than in any of 

 the above-named varieties of felspar ; but it ap]jears to mc that 

 it is by no means unphilosophical to suppose that a part of the 

 alkali in felspar mav be volatilized by a long continued heat. 



Secondh", both obsidian and pumice " give in fire exactly the 

 same products," and " both these fossils not only accompany 

 each other at Lipari, but likewise fiequently occur actually 

 blended." — Klaprotti's Analyses, vol iii. 



There is a circumstance relative to t'le origin of pumice, which 

 I own I cannot account for. It is this, that obsidian and lava 

 should contain pyroxene, while the analysis of pumice-stone does 

 not even indicate the presence of either lime or magnesia, both 

 of which were found in considerable quantity by Vauc[uelin in 

 his analyses of the pyroxene of ^Etna, and in the granular variety 

 of the same mineral, called coccolite. — Vide Ljicas Tableau des 

 Especes Minerales, part i. Paris 180S. 



The fact, that obsidian and pumice, though so different in 

 composition, give the same products by heat, may appear to mi- 

 htate against the above ideas ; but it seems probable that the 

 felspar only is fused in the experiment, and that the pyroxene is 

 merely enveloped in the paste formed by the former mineral, as, 

 if an attempt is made to fuse pounded granite, the felspar is vi- 

 trified, while the quartz remains in its pristine state enveloped 

 in it. 



B b 2 I shall 



