98 Cote—The Rhyolites of the County of Antrim ; witha Note on Bausite. 
They measure about ‘05 mm. from end to end, and resemble geniculated twins. 
The angle between their components can be measured in a number of examples, 
those being selected in which the two arms appear parallel to the plane of the 
section. The average of six measurements yields an angle of 123°, the lowest 
figure being 1143°, and the highest 130°. It is obvious that, in such small crystals, 
examples may be selected which are symmetrically disposed, but in which the 
angle of the twin is towards the observer, and the plane of the arms slightly 
oblique to that of the section. In such cases the apparent angle will always be 
larger than the true angle. The angle of the geniculated twins of rutile is 114° 
26’, and I do not hesitate to refer these abundant microlites to that mineral. 
Sometimes a second geniculation occurs in them, opposite in direction to the first, 
and the resulting crystal is somewhat Z-shaped. 
What has become, however, of the titanium oxide in the almost colourless 
layers of this rock? There can hardly be so striking a chemical differentiation 
between adjacent bands as to cause all the titanium to accumulate in the brown ones 
only. We are here in face of the many problems relating to the conditions which 
promote the growth of mineral species. Probably in the light-coloured layers— 
with the exception of that already noted, in which red rods actually occur—the 
segregation of the iron oxides as specks of magnetite has drawn the titanium into 
natural association with them, a titaniferous iron ore resulting. Had sufficient 
lime been present, and less iron, sphene might have been produced instead. But 
now follows a second problem. In what holocrystalline rock does such an 
abundance of rutile occur as is seen in the brown layers of this lava? What 
conditions have allowed the titanium to get so clear a start of other chemical 
constituents? I fear that we must leave the matter here, and return from 
speculation to the structural details of the rock. 
Perlitic structure occurs in both the brown and the lighter bands, and can be 
best seen in the latter when the light is thrown up obliquely. Mr. Watts* has 
observed a curious fact in sections of the obsidian of Sandy Braes, which is also well 
seen in the browner layers of the present banded specimen. The interior of some 
of the perlitic globules appears distinctly paler than the ground outside them, and 
occasionally a patch, more or less central, is actually colourless. In one case a 
globule extends into both a brown and an adjacent almost colourless band, and 
the clearer patch is even more translucent than the lighter of the two bands. 
Certainly, as Mr. Watts has remarked, some bleaching seems to have gone on. 
The same puzzle is likely to crop up in the examination of other perlites; but its 
explanation proves to be singularly simple. With a high power, the paler areas 
and their rim of darker glass cannot be brought into focus at the same time; we 
are looking down into a concavity in the surface of the section, the edges of which 
* Op. cit., Quart. Journ. Geol. Soc. London, vol.1., p. 370. 
