BY W. N. BENSON. 135 



the igneous rock can scarcely be doubted. The keratophyre, 

 when fresh, has a translucent white colour, and is divided up 

 into small portions, generally about two millimetres in diameter, 

 closely compacted together, and generally without any intervening- 

 matrix. A little pyrites is usually present. On weathering, a 

 white kaolinitic rock is produced, or a red or yellow ochreous 

 one. Towards the centre of the mass, and also to the north-east, 

 magnetite-keratophyre is developed. The passage from the one 

 rock into the other is a gradual one through many peculiar inter- 

 mediate types to be subsequently described. The increasing- 

 amount of magnetite may be regularly distributed, giving a 

 uniform grey rock, but more usually it is very irregular, pro- 

 ducing breccia-like or nodular rocks. Following these, are almost 

 pure magnetite-keratophyres with but few veinlets of felspathic 

 rock. Finally, in the centre of the ferruginous area, is a black, 

 heavy, and very slaggy magnetite-keratophyre, in which the 

 abundant irregular vesicles are never amygdaloidal, and are 

 nsually filled with calcite. A similar passage from purely 

 felspathic into richly magnetitic rock occurs on the summit of 

 the hillock. Between this and the former mass, is a long intru- 

 sion of quartz-dolerite, containing very little augite, and passing, 

 on the margin, into fine-grained, non-porphyritic rock, the com- 

 position of which approaches that of an augitic quartz-kerato- 

 phyre. North of the jasper-band, is a lenticular patch of very 

 vesicular red-brown quartz-magnetite-keratophyre. The vesicles 

 are round or slightly amygdaloidal, and the rock is not at all 



Along the zone of agglomerate to the north, other small occur- 

 rences of quartz-magnetite-keratophyre may be found, and the 

 line of contact between these and the agglomerate is very difficult 

 to define. 



At Silver Gully, a mile north of the Hyde's Creek complex, is 

 a second complex, of which a sketch-map is given (Fig.5). This 

 is only very roughly drawn to scale. Beside the agglomerate, is a 

 band of limestone forming a bold outcrop. To the east of this is a 

 narrow band of spilitic agglomerate, beyond which is a long intru- 

 sion of dolerite like that in the Hyde's Creek complex. There 



