CoLte—The Rhyolites of the County of Antrim; with a Note on Bauxite. 97 
Smeeth* has critically examined perlitic structure, with a special view to Mr. 
Watts’ conclusions, and declines to class together the cracks in the matrix and in 
the quartz, ‘‘save on an extremely general basis of classification.” He remains 
‘unconvinced that an originally lithoidal groundmass has been found exhibiting 
of itself perlitic structure.” Both Mr. Smeeth and Mr. Watts appear to use the 
word ‘lithoidal” in the sense of ‘‘ minutely holocrystalline.” If this is so, I 
agree fully with Mr. Smeeth’s remarks; but it must be remembered that the 
‘“lithoidal” rocks of old French writers often contain, as above noted, a good 
residuum of glass, so that even stony-looking rocks may prove in section to be 
hemicrystalline and occasionally perlitic. 
On Sandy Braes, as usual, the perlitic structure is more and more imperfect in 
proportion as the rock is more lithoidal, and it is conspicuously absent in those 
lavas which are nearly or actually holocrystalline. On the other hand, it is only 
fair to Mr. Watts to state that the curvilinear cracks in quartz occur in their best 
development in those rocks which are most admirably perlitic. Similar causes 
have produced the two types of cracks in the erystals and the glass respectively.t 
A beautifully fluidal and banded perlite was picked out of the surface-sand in 
the diggings south of Sandy Braes. In the specimen, the glass consists of alternating 
purple-grey and opaque cream-coloured bands, the flow round the porphyritic 
crystals being most effectively seen. In microscopic section, the creamy bands 
become grey and almost colourless, while the others are a strong yellow-brown 
(Pl. IV., fig. 4). The almost colourless bands are the more lithoidal, as may 
be seen between crossed nicols; but they contain at the same time a larger 
quantity of pure glass, the primitive crystallites having withdrawn themselves to 
form a multitude of little greenish or colourless prisms, and specks which are 
probably magnetite. One of these lighter layers is exceptional, the prominent 
microlites being dark red rods, on which a number of colourless platy crystallites, 
of high refractive index, lie transfixed. The material in the brown layers is less 
differentiated within itself, and the colouring matter in the more translucent parts 
remains unresolved, even on the thin edges of the section and with a power 
magnifying 400 diameters. In the more dusky of these brown layers, ‘‘ cumulites,” 
aggregates of globulites, can be distinguished as causing the cloudy effect. 
Crystallites occur in the brown bands, similar to those in the lighter ones, but 
more sparsely developed ; and in addition there are numerous and remarkable red- 
brown microlites, in the form of comparatively long rods bent about the middle. 
* A Perlitic Pitchstone from the Tweed River, New South Wales, with remarks on the so-called 
Perlitic Structure in Quartz,” Journ. Roy. Soc. of N.S. Wales, vol. xxviii. (1894), p. 306. 
} Since the reading of the present paper, Mr. Watts has published (Geol. Mag., 1896, p. 15) a 
review of Mr. Smeeth’s observations, which does not, I think, affect the remarks that I have made 
above. 
