Vol.57.] 



OF COKFNDUM AS A COXTACT-MINERAL. 



187 



A minute mosaic of colourless grains, with low refractive index 

 and weak interference-colours, fills up the background; and in many- 

 cases these show the multiple twinning of plagioclase. Twinning 

 on the albite and pericline plans is found : in one case also, appa- 

 rently Carlsbad twinning occurred. Among others, the following 

 nearly symmetrical extinction-angles were measured from (pre- 

 sumably) albite twin-lamellae : — 12 • 11, 12| • 13. 



Some few individuals which appear untwinned, and extinguish 

 approximately parallel to their greatest length, may be orthoclase. 



It is difficult to . . '' . 



be certain about ^'^S- ^- — Individual of corundum seen in a thin 

 the occurrence of section of one of the included fragments. 



quartz. I feel sure, oooi 



however, that very 

 little or none is pre- 

 sent, for close ex- 

 amination of many 

 tiny grains in the 

 colourless mosaic 

 which suggested 

 possible quartz, re- 

 vealed faint multi- 

 ple twinning, or a 

 portion of a biaxial 

 interference - figure 

 was obtained. The 

 numerous inclu - 

 sions of the colour- [The crystal is cut at right angles to the basal plane, 



less minerals ren- 

 der their exami- 

 nation somewhat 

 troublesome. 



Iron - ore is 

 abundant in scat- 

 tered grains, and 



which is slightly stepped owing to alternation with 

 a rhombohedral face. The faint horizontal shading 

 indicates approximately the distribution of the pale 

 blue colour. The sharp rhombohedral cleavage- 

 cracks are drawn, as if seen between crossed nicols ; 

 in ordinary light they are only with difficulty 

 distinguishable. The few inclusions consist of 

 iron-ore, a little sillimanite, a grain or two of 

 zircon (?), and gas-inclusions, X 38"6.] 



also as inclusions in 



the other minerals, including corundum. 



Sillimanite is locally abundant, elsewhere scarce or perhaps 

 absent. Where abundant, it forms sheaves and aggregates of much 

 elongated colourless crystals. The characteristic transverse parting 

 is seen, the segments being often separated by a portion of the 

 enclosing mineral. 



Green spinel occurs in irregular, rather small grains, or granular 

 aggregates. 



A few fairly well-shaped crystals of zircon were noticed. 



In a slice passing through the actual contact, there appears to be 

 some injection of felspathic material. Prof. Barrois noticed that some 

 of the inclusions were penetrated by very fine veins of the enclosing 

 granite (especiallv rich in hornblende). In view of the considerable 



o2 



