F.vuQuriARSON. — The Platinum Gravels of Orejniki. 



465 



usually elongated in the direction of the pinacoid faces. No twinning 

 has been seen. The hornblende is often much altered. The more usual 

 alteration consists in the development round the borders of the plate of a 

 secondary outgrowth of the mineral, causing an apparent extension of the 

 plate. Another type not uncommon in the sections examined consists 

 in a large development of chlorite, which is distinctly pleochroic, green in 

 colour, and very ragged in outline. Now and again calcite and magnetite 

 are formed in small quantities. 



The (6) variety consists of plagioclase, augite, and hornblende. The 

 plagioclase has characters exactly similar to those in the [a) type ; the 

 hornblende also much the same as before. The augite is rather peculiar. 

 In some cases it seems to take the form of diallage, and with those cases in 

 which it occurs to a large extent it would lead one to call the rock not a 

 diorite, but either an augite-diorite or a gabbro. The diallage occurs in 

 large plates with a more or less well-marked schiller structure formed by 

 rows of inclusions, an extinction-angle of nearly 40°, strong orthopinacoidal 

 cleavage, and with no pleochroism. The plates are also usually colourless, 

 and the inclusions are often very numerous. Augite as distinguished from 

 diallage also occurs. This is without regular outlines, with an extinction- 

 angle of 35° (approximately), and quite colourless in section. Alteration, 

 a^i with the hornblende, is sometimes pronounced. The usual type of 

 alteration is the development of uralite. This begins at the border of the 

 plate, gradually extending inwards, and frequently a ragged scrap of augite 

 is enclosed almost entirely by uralite. The latter has only a very faint 

 pleochroism. In another type of alteration chlorite is formed, and even 

 calcite, just as in the case of the {a) type of rock. 



The structure of the rocks presents characters strongly suggestive of 

 dynamic action. On first looking at a section with crossed nicols one 

 cannot help being struck with the " crushed " appearance of the rock. 

 The whole section seems to be composed of a mosaic of small plates of 

 feldspar and ferro-magnesian mineral placed in a curiously disorganized 

 manner. The appearance is exactly the same as the granulitization seen 

 sometim.es in granitic masses, and is probably due to the same cause — 

 i.e., crushing of the rock-mass after consolidation. 



Sometimes there is an appearance very like the grounduiass seen in 

 quartz-porphyries that are holoerystalline — namely, a very small mosaic — in 

 this case consisting of altered hornblende or augite and little plagioclase plates. 



An analysis of the rock runs as follows : — p^^^, q^^^ 



SiO, .. .. .. .. .. 47-40 " 



FeO 



Al.,0., 



Cab " 



MgO 



K2O 



NaaO 



H2O 



^6303 



MnOo 



TiO," 



Note. — In this analysis the CaO was determined volumetrically with 



KMnOj standard solution. This was the 

 of the coral limestone at Funafuti, and 



6-60 



18-17 



12-23 



7-17 



0-20 



2-75 



0-75 



5-42 



Trace. 



Trace. 



method followed by the analysts 

 it certainly seems neater and just 



as accurate^as the other way. On fusion of the powder with the fluxes 



