104 ME. J. PAKKIlSrSON O'N THE PIEOMERIDES [Feb. 189S, 



sides, and when at one point the irregular mararaillated edge of the 

 nodular band has given rise to an indentation which is included 

 in the section, the radial structure is seen to keep normal to it 

 also. There results, of course, a line analogous to that formed by 

 the ingrowing microliths in the devitrified glass described by 

 Prof. Bonney ^ ; and it seems fairly obvious that such a surface of 

 discontinuity must have represented a factor of importance in 

 determining the position of rupture under the strains to which the 

 rock has been subjected. 



In the second specimen of the series, the constrictions or inden- 

 tations which were but slightly developed in the first become more 

 marked, giving a corrugated appeaiance, with a very fairly circular 

 section, to the band. It may, in fact, be described as a series of 

 spheres, each of which has been ground down in two parallel 

 planes and the ground edge of one applied to the ground edge of 

 another. The band itself is dark purple, and harder than the 

 surrounding rock, breaking with a clean fracture ; it is, in a word, 

 identical with the felstone forming tt > pyromerides. 



A third specimen essentially resembles the last, but shows a 

 mammillated appearance which is rather characteristic, the nodules 

 appearing as though covered with imperfectly differentiated buds, 

 while the larger lumps, from which those have as it were sprung, 

 may be seen clearly outlined, though not clearly differentiated the one 

 from the other. A thin section of this specimen has been prepared, 

 but it presents no points of especial interest. The slide, as seen by 

 the unaided eye, resembles two mutually-interfering pyromerides, 

 and the microscope shows no points of difference therefrom, either 

 with ordinary or with polarized light (PI. YIII, fig. 1). Each 

 pyromeride contains at its centre a roughly circular patch, rather 

 darker and with a more confused type of depolarization than that 

 shown by the surrounding matter. Occasionally tlie constituents 

 of these patches are arranged so as to leave irregular and clearer 

 areas, which polarize uniformly, in an opaque setting which is much 

 more inert between crossed nicols. This structure is by no means 

 uncommon in pyromerides, though varying much in the extent to 

 which it is developed. In relation with the two constituent 

 pyromerides are two crescentic areas of quartz, 10 times as long as 

 broad, very nearly concentric with the respective edges, and in both 

 cases at the same distance from them. The curvature of the two 

 arcs is, of course, not identical, since the diameters of the component 

 pyromerides are not the same. 



Continuing the series, it is seen that the next step to this is an 

 irregularly oval pyromeride with a suggestion of tapering at the 

 ends, and irregularly mammillated. Stages can be found easily 

 enough in these cliffs, in which one egg-shaped pyromeride has 

 been apparently constricted off from its neighbours : they themselves 

 being perhaps not so clearly defined, but occurring in a little group of 

 three or four mutually-interfering individuals. It is interesting 



^ Quart. Journ. Geol. Soc. vol. xli (1885) Proc. p. 92, also Herman & Rutley, 

 Proc. Roy. Soc. yol. xxxix (1885) pi. i, fig. 1, pi. iii, fig. 5 & p. 90. 



