Marshall. — Geology of Tahiti. 



371 



To some extent at least the great mineralogical difference that has 

 been mentioned can be explained. It appears that hornblende and mica 

 do not crystallize in an igneous magma unless some mineralizer is present. 

 Thus Doelter says, " Was die hornblende anbelaxgt, so wird vielfach die 

 Notwendigkeit von Wasser und hoheren Druck angenommen. Keinenfalls ist 

 ihre Bildung ohne Mitwirkung von Mineralisatoren moglich, ebensowenig wie 

 bei Glimmer . . . Bezuglich der Hornblende hat Becke der Einfluss des 

 Druckes aufihre Bildung studiert and scheint jedenfdlls wenigstens eine geringe 

 Druckzunahme ihre Entstehung zu fordern."* 



At the time of an eruption of magma the water, and probably any other 

 mineralizer, would escape, and there would then be no chance of hornblende 

 forming. A new molecular arrangement within the magma would be 

 the result. This might well end in the crystallization of the minerals of 

 a normal basalt. It is noticeable that, while theralites and nephelinitoid 

 monzonites are apparently the commonest types of plutonic rocks, basalts 

 are by far the commonest types of volcanic rocks. 



Lacroix has already shown that a clear relationship of a chemical nature 

 exists between his monzonite nephelinique a biotite and the hauynophyre. 

 It can be clearly seen from the accompanying table that the syenite men- 

 tioned in these pages is a chemical representative of the phonolite described 

 by Lacroix as coming from Vairao. The peridotite also obviously belongs 

 to the same chemical class as the picrite he has described. The table of 

 analyses given here is taken from his paper, but the analyses made by me 

 of the syenite and peridotite are added to it. 



Plutonic Rocks. 



A. Syenite with pyrite and a little biotite. Analysis, P. Marshall. 



B. Syenite nephelinique a biotite. 



C. Syenite nephelinique a ampliibole. I 



D. Monzonite nephelinique. - Analysis, M. Pisani. 



E. Gabbro nephelinique. 



F. Gabbro essexitique. 



G. Peridotite (wehrlite). Analysis, P. Marshall. 



B, C, D, E, and F quoted from Lacroix, " Roches alcalines de .Tahiti,' 

 Bull. Soc. C4eol. de France, 4 e serie, t. x, 1910, p. 121. 



* C. Doelter, " Petrogenesis, Braunschweig," 1906, p. 146. 



