106 ME. J. PAEKINSON ON THE PYR03IEEIDES [Feb. 1 898, 



definite dovetailing edges allow of no blurred appearance as the 

 stage is rotated. Accordingly several structures can be seen 

 normally in these nodules, formed probably in the following order : — 

 (i) A flow-structure marked by black microliths, (ii) a radial 

 structure, and (iii) the patchy structure seen only between crossed 

 nicols. 



In many cases part of the original surrounding matrix is found 

 still adherent, after the pyromeride has been detached. This 

 frequently is much less opaque in a thin section than the nodule 

 itself. It contains no microliths, but has much of a secondary 

 green mineral developed, which, in addition to being spread 

 irregularly in patches, also defines lines and curves more or less 

 concentric with each other ; these seem certainly to be perlitic 

 cracks. I^ow, as a rule the rock is microcrystalline. 



It is common to find very small spherulites, which have sprung 

 from the surface of the nodule, forming growths which resemble a 

 mushroom without its stalk. Apparently the surface of the nodule 

 has been one of discontinuity and has favoured these growths, as in 

 the partially-fused sheets of glass described by Prof. Bonney.i 



A further characteristic of structural importance remains to be 

 considered in the frequent presence of a crescent of quartz, visible 

 usually towards the edge of the nodule when it is broken open, 

 Por the sake of simplicity, it may be well to take a definite case 

 from such an example as that just described. 



The pyromeride measures 2x1^ inches; the quartz crescent, 

 about ^ inch from the top, 1^- x '075 inch. It tapers slightly 

 at either end, and has a zoned appearance suggestive of infil- 

 tration. At about I inch from one extremity the crescent turns 

 upward^ towards the edge of the nodule; at the other, it 

 follows with some exactness a change in the curvature of the 

 peripher}'-, and at this point has internally to it other less well- 

 developed crescents of quartz. At the other end also it can be 

 seen in a thin section to bend slightly upward, although not so 

 markedly as at the first. These are analogous to the ' points de 

 rebroussement,'^ described by Delesse. 



In a section taken at right angles to the longer axis, the crescent 

 is seen to be continued for ^ inch, and here also, near the point 

 where it dies out, its concave and not its convex side is turned 

 outward. The general appearance suggests that the contraction 

 of the nodule has resulted in a crack, and it may be well to see 

 what evidence there is for this in other cases. 



One instance of these quartz-filled arcs has been given already in 

 discussing the microscopic appearance of the band-like aggregation 

 of pyromerides in an earlier joart. Some points then mentioned, 

 such as each crescent having the same thickness of external wall, 

 although the diameters of the component nodules are different, 



^ Quart. Journ. Geol. Soc. vol. xli (1885) Pi'oc. p. 91. 

 ^ The change of direc<"ion at this point is about 150°. 

 2 Mem. Soc, geol. France, ser. 2, vol, iv (1852) pt, ii, p, 314. 



