520 THE CONTINENTAL ORIGIN OF FIJI, II., 



The chemical anal3'sis of this rock made by Messrs. Stoddard 



and Mawson at the Universit}' of Sydney, by kind permission of 

 Professor Liversidge, is as follows: — 



SiOo 58-92 



AUOy iG-95 



re.,0,, 3-33 



Feb 3-69 



MgO 0-40 



CaO 6-22 



Na.,0 4-99 



K.b 3-08 



HoOatllO^^ 1-09 



HoO on io-nition ... .. 1-27 



Total 99-94 



This indicates a somewhat basic andesite. 



Pyroxene Andesite (from coarse volcanic breccia of 

 Voma, Upper Waidina River). 



Macroscopic characters. — The breccia, from which the rock here 

 described was obtained, forms an enormous outcrop. It rises 

 from the river-level in great cliffs which must be quite 1000 feet- 

 high. These cliffs limit the view in nearly all directions, the 

 greatest distance visible being somewhat over a mile. The 

 boulders range up to 4 or 5 feet in diameter, and are angular to 

 subangular in shape. The groundmass is greyish in colour, and 

 consists of comminuted fragments of rock similar to that com- 

 posing the boulders. The latter in hand specimen is of a rather 

 light blue-grey colour. It is minutel}^ vesicular, the cavities 

 being irregular in shape, and up to about 2 mm. in diameter. 

 They are coated inside with a thin film of bluish-white substance 

 which does not effervesce with acids. These give the rock a 

 speckled appearance. There are numerous glass}-^ striated felspars 

 up to 3 or 4 mm. in length, with less abundant and less obvious 

 small cr3^stals of brownish translucent pyroxene. A certain 

 amount of magnetite can also be detected. The presence of the 



