Rer. T. G. Bonney — Pitchstones and Felsites of Ar ran. 503 
will show that though this may often be the case, the mechanical 
forces produced by contraction have also sometimes an influence 
upon it, and may even detennine whetber or not this structure sliall 
be set up. 
The first case which I shall quote is not from Arran, but is ex- 
liibited in a slide of liparite from the Lipari Islands. It is a por- 
phyritic rock containing many crystals of felspar, with a little quartz, 
biotite, and iron-peroxide. An ordinary lens shows the dull white 
ground-mass to be traversed by a series of more transparent lines 
dividing it up into polygons, so that it looks like a miniature represen- 
tation of a columnar-jointed surface. Inside the polygons, the rock 
exhibits a faint radial structure. Under the microscope these cracks 
appear to have been true minute divisional surfaces, though they now 
seem to be closed by a generally paler mineral deposit, which is 
dark between crossed prisms. Sometimes also we see a thin dark 
line, bordered on each side by a clear space. The inferior of the 
polygons commonly exhibits very clearly the usual fibrous radial 
structure ; 1 but in some of the smaller it is very ill-defined. In 
sorne cases we note a kind of circle inscribed in the polygon, within 
which the radial structure is more marked ; generally there is no 
nucleus visible in the spheroids, though now and then they seem 
to become rather more crystalline on approaching the centre ; nor do 
they appear to have any very immediate relations with the larger 
embedded crystals (compare Fig. 3). 
Passing uext to a perlitic pitchstone from Meissen, we observe 
that it exhibits a flow-like structure, indicated by wavy bands (in 
outline like some cirrus clouds), produced by innumerable very 
minute opaque microliths. Through these the perlitic structure 
cuts with perfect independence. There are no spherulites to be seen 
in the slide. A perlite, also, from Glashütte (Schemnitz), shows very. 
well this wavy flow-structure ; but here the bands are so conspicuous 
and clearly marked that the rock looks like a miniature model of a 
contorted gneiss. A few spherulites occur in this, which in most 
cases, though not in all, seem connected with the larger crystals 
occurring in the matrix. 
The great pitchstone vein on the Corriegills shore exhibits, as bas 
been said, faint indications of a banded structure. Under the micro- 
scope it appears full of a microlithic dust with larger belonites 
singly or in groups, surrounded by slightly clearer spaces ; a good 
many of these belonites are roughly parallel. The smaller vein has also 
indications of banding, though slighter, with a similar microscopic 
structure : but the clearer spaces are larger, and themselves indicate 
a kind of banding. Tnere are one or two spherulites surrounded by a 
clearer ring. More distinct bands are seen in the lower pitchstone vein 
of Dundhu, and some of the dykes of the Tormore shore, the micro- 
scopic structure of which is excellently depicted by Mr. S. Allport, 2 
who says (p. 6), “The rock is simply a continuous homogeneous 
glass, in which the devitrified or crystallized particles have arranged 
1 See Mr. S. Allport, Geol. Mag. Dec. I. Yol. IX. p. 541. 
2 Geoe. Mag. Lee. I. Vol. IX. Pi. I. Fig. 2. 
