410 EEPORT— 1888. 



The first is from a dyke east of Mynydd Mechell (225) (see fig. 20). In 

 this there are no proper insets, but here and there are little collections of 

 epidote crystals — sometimes almost isolated and sometimes attaining the 

 magnitude of a fair-sized spherule — but they are not radiate, nor regular in 

 outline. The groundmass, as seen by the paraboloid, looks like an almost 

 uniform glass, but dim and shadowy outlines separating very feebly 

 dotted spaces indicate a subdivision into larger elements. When seen 

 between crossed Nicols these larger elements are still conspicuous as a 

 kind of substratum, or earlier stage of division ; but they are greatly 

 interfered with by the development, with sutural boundaries, of numerous 

 smaller crystals, whose long axes are in general irregularly arranged, but 

 ■which in many spots radiate from a point, like ill-formed spherulites, or 

 from the circumference of the epidote nests. This combination of larger 

 elements with smaller spherulitically arranged ones is so peculiar that if 

 it is met with elsewhere it should receive a name, such as macro-felsitic 

 structure. In a second dyke from the same district (226) the ground- 

 mass is microgranitic, and there are no insets or epidote nests, but it is 

 crowded with large spherulites of the ordinary fibrous radiating t3rpe. 

 These have occasionally successional rings ; where most crowded the 

 interspaces are filled with larger elements. Its present state has there- 

 fore been reached by at least two stages. The rock at a large boss at 

 Tygoch-wyfsa, east of Amlwch (230), is a typical microgram te, whose 

 insets are idiomorphic quartz and twinned orthoclase, which are often 

 surrounded by a spherulitic border. The most beautiful of all these rocks, 

 however, is that of Hafod-onen, south of Amlwch (228) (see fig. 21), 

 marked on the Survey map as ' greenstone.' Its original structure is rather 

 obscured as seen between crossed Nicols, but with the paraboloid it is 

 seen to be truly porphyritic, all the insets being felspar and most of them 

 idiomorphic, and some speckled. In the groundmass there is a little 

 microgranite, but the characteristic structure is that of a beautiful 

 granophyre. The elements composing this are very minute, and are 

 arranged with the longer diameters of several in succession in one 

 direction, so as to form a lacework pattern. The difference between the 

 materials of the elements cannot be demonstrated between crossed Nicols, 

 but by oblique light there is seen to be a difference in the refractive index. 

 The structure is therefore that of a true micropegmatite. In many cases 

 the network starts from the sides of a small rectangular inset and grows 

 out in bushes, forming an hour-glass pattern in the mass, which is 

 bounded on all sides by similar growth from other crystals. They thus 

 form the pseudo-spherulites of Rosenbusch, with this difference only, that 

 their outlines are regulated by the interference of neighbouring growths 

 and by the size and shape of the originating crystal. 



Another group of rocks consists of those which weather spheroidally, 

 and all of which have a more or less common structure. One of the 

 simplest is largely developed among the volcanic rocks of Careg Gwladys 

 (206). In this tlaere is an abundant base, which remains dark between 

 crossed Nicols, but it is scarcely a glass, to judge by its dusty appearance in 

 ordinary light, but it may be microfelsitic — there are a few minute polar- 

 ising specks in it. In this groundmass are scattered numerous microliths 

 of a lath-shaped plagioclase, which, from its low extinction angle, is pro- 

 bably oligoclase. These lie in every direction and interlace with each other 

 without any regularity. We have thus the absolute reverse of orientation, 

 seeing we start with axial crystals. To distinguish this we may compare 



