THE VOLCANOES AND ROCKS OF PANTELLERIA 



HENRY S. WASHINGTON 

 Geophysical Laboratory, Carnegie Institution of Washington 



PART III 



PETROLOGY 



Chemical characters. — ^All of my analyses of the Pantelleria rocks, 

 with their norms, are given in Tables I and II. Those of Foerstner 

 are omitted for reasons to be given later. They are sharply 

 divided into a large group with high silica, varying from 63.30 

 to 72 . 21, and a smaller, the basalts, with about 46 per cent of silica. 

 So far as known, from Foerstner' s publications, my own observa- 

 tions, and Butler's collection, there are no rocks of intermediate 

 composition. 



Considering the larger group first, rather high alumina is found 

 only in two or three cases, this constituent being in most of the rocks 

 unusually low for rocks of such moderately high silicity — if the 

 coining of a new word on the analogy of acidity and basicity be 

 permissible. On the other hand, the iron oxides are distinctly 

 high, especially in the pantellerites, going hand in hand with low 

 alumina. Magnesia is uniformly low, as is lime, the latter rarely 

 attaining more than i per cent. The alkalies are distinctly 

 high, with soda dominant over potash, the latter being remarkably 

 uniform. Titanium is remarkably high for rocks of this silicity 

 and alkahnity. Phosphorus and manganese are both rather above 

 the normal, but zirconia is low for sodic rocks, and only traces of 

 barium are present. Traces of nickel seem to be commonly present, 

 but tests failed to reveal the copper reported by Foerstner. 



The basalts are very uniform in composition and show only 

 one especially noteworthy character — the high titanium content.^ 

 Magnesia is not high for basalts of this general composition (camp- 

 tonose) and alumina is distinctly low, much lower indeed than the 



^H. S. Washington, Q.J.G.S., LXIII (1907), 76. 



16 



