108 



MR. J. PAEKINSOX ON THE PYROMEEIDES [Feb. 1 898, 



themselves present no great peculiarity. Overlying these comes a 

 spherulitic rock, identical with, or at all events closely resembling, 

 that recently analysed by Mr. Hyndman.^ Slightly farther up 

 these spherulites (or pyromerides) become larger, the size of a small 

 cherry, often showing concentric and radial structure in a greater 

 or lesser degree. There is often a crescent- shaped mass of quartz 

 present, although solid specimens without conspicuous radial struc- 

 ture are not uncommon. The rock in which they are set is softer 

 than they themselves are, and pale green in colour. 



Pig. 1. — Diagram {see explanation below). 



a = central quartz-mass. 



b = crack nearly ^ inch broad. 



c = ill-developed pyromeride, passing into 

 d = mass of rock. 



A few yards farther on the rock becomes of a streaky nature ; 

 that isj a rock which, although plainly showing flow-structure, has 

 yet the flow-bands with irregular boundaries on the fractured face, 

 and these, instead of remaining strictly parallel, often join and mingle 

 with each other : in other words, a rock part of which has not 

 been particularly fluid. 



The bands and pyromerides are formed partly of a purplish material, 

 partly of a material similar except for the absence of ferrite, and con- 

 sequently of colour ; while the rest of the rock is of a yellowish-brown 

 tint inclining to green, the difference in colour causing the relations of 

 the two principal constituents to be conspicuous. The tendency to 

 segregation into spherical forms, which the more ferruginous matter 

 shows, studs the rock with pyromerides seemingly in all stages of per- 

 fection, an effect partly due to the absence of the ferrite just 

 mentioned in much of the pyromeride-forming matter, the whole of 

 this not being at once obvious. With these pyromerides are found 

 many gas-vesicles. These ' spheres ' occur in patches often 

 suggestive of a sort of flow-brecciation ; an irregular but band- 

 shaped patch where ' spheres ' are abundant, may be succeeded by a 



1 Geol. Mag. 1896, p. 36.5. 



