Vol. 65.] AND MAGNESIAS' EOCKS OF NEW ZEALAND. 369 



(4) Auhunga. — A somewhat opaque variety, with the green colour of 



Kawakawa and the opacity of Inanga. 



(5) Totoweka (Weka's blood). — A variety of Kawakawa containing 



stains and streaks of red iron oxide. Hence its name. 



(6) Raukaraka. — Named from its resemblance in colour to the leaf 



(rau) of the Karaka {Corynocarpus laevigata). An olive-coloured 

 streaked or cloudy variety, often with a yellowish tinge. 



To the foregoing list might be added tangiwai or koko- 

 tangiwai, the bowenite of Milford Sound. The word tangiwai 

 means ' tearwater,' and was given to this stone on account of the 

 delicate reflections and the appearance of drops of falling water 

 occasionally seen in polished specimens by transmitted light. The 

 bowenite has been already described. 



Mr. Chapman has applied the Radde colour-scale and nomen- 

 clature to denote the different shades [13] ; but, owing to the fact 

 that almost every conceivable variation of tint and shade may be 

 seen, even in a single specimen, the scale is not of great practical 

 utility. 



Causes of the variations in colour and appearance. — 

 The essential pigment of the mineral appears to be ferrous sili- 

 cate. A series of analyses made by me of specimens ranging 

 progressively in colour from the pearly Inanga to the deep-green 

 Kawakawa showed a proportionate increase in the percentage 

 of combined ferrous oxide (see later) ; there can be consequently 

 little doubt that the essential colouring-matter is ferrous silicate. 

 The composition, in fact, varies between that of tremolite and 

 actinolite, although most of the specimens are actinolitic in com- 

 position. Infiltrated oxides of manganese and iron account, for 

 the most part, for the black opaque spots, secretions, and markings 

 which are frequently present : these, indeed, largely detract from 

 the appearance of such specimens for jewellers' purposes. 



Thickly clouded and variegated specimens, exhibiting varying 

 shades of green, show corresponding irregularity in the percentage 

 of ferrous oxide ; the variegation is, therefore, most likely due to 

 local secretion of this constituent in its combinations. Alternate 

 dark and light bands in the stone are probably a result of original 

 differences of chemical composition. 



Again, streaks in the stone may accompany cracks and flaws, 

 with the separation of iron oxide in the cracks and its infiltration 

 into the surrounding mass, giving dark shades of colour. The 

 red stains of the local variety, totoweka, are likewise due to 

 infiltrated ferric oxide. 



Finally, flaws and cracks impart to the surface a mottled appear- 

 ance, with white or opaque splashes of colour, due to reflections 

 from the surfaces of fracture. Thus most of these variations are 

 due to accidental and minor causes, and do not, so far as can be 

 seen, indicate in any case particular modes of origin. 



