ON THE OLDER ROCKS OF ANGLESEY. 405 



neiglibourliood of Gaerwen, essentially of the same nature as tlaat at 

 Craig-yr-allor, bat more foliated throughout. The closest resemblance 

 is in the so-called 'gneiss ' by Holland Arms (175), which has the same 

 uniformly distributed small elements of felspar and hornblende and 

 numerous minute sphenes. Some few of the hornblende elements have 

 small corners which present the characters of a pyroxene ; but their 

 long axes are all in the same direction, and they are not defoi'med. In 

 every other case examined the elements have been more or less pulled out 

 into lenticles, whereby their aspect is greatly changed. The first stage 

 of this process and its results are beautifully seen in the rock at Gaerwen 

 Windmill (173). Here the hornblende has become rather fibrous, but is 

 otherwise little changed except in shape ; but the felspar, assuming it to 

 have been originally present, is now in the form of dusty, not very trans- 

 parent lenticles, which have the optical characters of epidote. These 

 ha^e parallel sides for some distance, and then terminate gradually. 

 The quartz which was present is found in microspectrally polarised bands. 

 There are also the usual sphenes and a little pyrites. Some later veins 

 in this rock are of interest, being composed of clear felspar, with needles 

 of actinolite (?) radiating from the ends of the hornblende layers. In a rock 

 at the most northern end (182) we find a second stage. Here the same 

 elements of hornblende, epidote, and sphene can be seen to be present, 

 but the rock is now crowded with mylonitic lines, between which the 

 epidote is ground to powder, and only a few recognisable fragments of 

 hornblende are left. It is strange that this shearing should not have 

 been followed by any chemical changes, especially as there are trans- 

 verse cracks filled with calcite, and a very low polarising mineral, per- 

 haps serpentine. In this case the strain has acted so as to intensify 

 the foliation ; but in a rock from the wood north of Y-graig (179) it has 

 broken it up, so that orientation can only be observed in small frag- 

 ments. It is remarkable that no epidote can be observed in this rock, the 

 felspathic element having broken np and become microspectral or mosaic. 



The interpretation of some other rocks of larger elements will now be 

 easier. There is a banded rock composing the hill north of Y-graig 

 (178) in which dark-green bands alternate with white ones. This 

 corresponds on a larger scale with the rock at the windmill, the green 

 bands being elongated hornblende, and the white ones decomposed epidote 

 mnch broken up. The rock at Ysgubor Llwyd (180) is of still larger 

 elements, difficult to identify : some are certainly hornblende, and others 

 epidote and felspar, but there are large porphyritic crystals with strong 

 cleavages in one direction, and others occasionally cutting them at 87^°, 

 and the extinction makes an angle of 14° with the first. It may be a 

 felspar. Another is colourless and non-dichroic, in long flakes like mica, 

 with cleavages in the same direction, and the extinction making an angle 

 of 18° with these. It may be a pyroxene. The rock has been much 

 broken. A third, to the north of Holland Arms (181), is almost entirely 

 hornblende, in large frayed-out crystals with spectral polarisation, a 

 phenomenon rare in this mineral. A few narrow lenticles of epidote, and 

 crystals of sphene bring it into the same group as the rest. 



The rocks in the narrow band which runs up to Pentreath can be at 

 once compared with and understood by the aid of that at the northern 

 end of the ' gneiss ' area. One, at the crossing of the Beaumaris road 

 (183), which is entirely cataclastic, shows some relics of hornblende and 

 epidote, but the intervals are filled with calcite. In the rocks nearer 



