J. V. Elisden—Igneous Rocks of Lleyn, N. Wales. 305 
the northern flank of Mynydd-y-Graig, about midway between these 
last-named localities. In this rock plagioclase is abundant, but 
rather decomposed. The extinction angle is rather large, corre- 
sponding to anorthite. Some crystals look like orthoclase twinned 
on the Carlsbad type, but are shown by their symmetrical extinction 
of about 53° to the trace of the twinning plane to be plagioclase. 
The augite is sometimes tolerably fresh, extinguishing perfectly, 
but other crystals show aggregate polarisation. It is often inter- 
penetrated by felspar, and the margins are developed into strongly 
dichroic brown hornblende. A brownish granular substance is often 
arranged in bands across the cleavage directions, and sometimes 
there are inclusions of hornblende. There is also a good deal of 
fibrous yellowish green viriditic substance, resulting probably from 
the decomposition of augite and hornblende. In one case lines of 
opaque granular decomposition product intersect at the characteristic 
prism angle of augite. 
At Treheli the same general characteristics are noticed. Under 
the microscope the plagioclase is seen to be abundant, but rather 
decomposed, Where polarisation colours are still given, the ex- 
tinctions are apparently not so high as in the last-mentioned speci- 
men. ‘The augite is also not so fresh: it occurs in irregular masses, 
often developed into hornblende. Some crystals have a central 
patch of decomposition product, giving aggregate polarisation, the 
margins extinguishing perfectly. In other cases the margins are 
decomposed and the centre fresh. Large masses of a clear pale 
yellow substance, with granular patches here and there, show between 
crossed Nicols a central dark patch, changing but little on rotation, 
surrounded by densely interlacing fibrous crystals, with brilliant 
ageregate polarisation and feebly pleochroic. ‘These fibrous, tufted 
crystals occur chiefly at the junction of hornblende and augite. The 
hornblende is much less altered than the augite, many of the viridite 
patches being fringed by perfectly fresh-looking hornblende. 
Original magnetite seems scarce, but lines of secondary oxide of 
iron often mark the boundaries of augite crystals, the interior being 
occupied by viridite. 
North of the road near Tyganol-bwlch-y-rhiw, we begin to 
approach the so-called Rhos Hirwaun syenite, and here the rock 
undergoes considerable change in character. The felspar is here 
very opaque, and often encloses green microliths. The augite seems 
to be nearly all converted into viridite, which is interpenetrated by 
the felspars. Even the apparently unaltered fragments give agere- 
gate polarisation. Quartz is fairly abundant in crystals and irregular 
grains, and generally contains enclosures of apatite needles, as well 
as cloudy patches of fluid and gas cavities. No hornblende can be 
detected; but there are a few ill-defined flakes of brown mica. 
There is a fair quantity of magnetite, and some ilmenite, partly 
converted into leucoxene. This rock appears to be a quartz-diahase, 
but a comparison would be interesting with the diabase of Castell 
Carron, about two miles north of this point, in which Tawney recog- 
nized quartz of secondary origin. 
DECADE III.—VOL. V.—NO. VII. 20 
