70 H. J. JOHNSTON-LAVIS ON THE GEOLOGY 
of well-rounded pumice, each fragment having a yellow dusty 
covering. The fragments range from the size of a hazel-nut down 
to the finest dust. A band on any one given horizon is composed 
of fragments of uniform magnitude, and may be followed by one 
composed of much smaller or larger pieces. There are amongst the 
coarser beds two or three intervening bands of fine ashy matter. 
These, although only ranging from one to two centimetres in thick- 
ness, may be traced over very large areas, showing that the process 
by which they were brought into their present position was general, 
and not local as in the case of alluvial deposition. 
The internal characters of the pumice are pretty distinctive. The 
structure is of a light spongy nature, with many large compound 
cellular cavities. The matrix is of a grey colour, usually with a 
yellowish or greenish tinge, which is broken rather abundantly by 
crystals of augite, often large, besides a few of sanidine and biotite. 
Mixed with the pumice are exceedingly few small lapilli, the 
whole bed seeming to be composed of primary matter. Crystals of 
pyroxene, loose and often broken, sometimes reaching a centimetre 
in length, are found loose with the pumice. They are similar to 
the crystals found in the rock itself, and are therefore probably only 
the same detached. 
Microscopical Exammation.—The first mineral that attracts one’s 
notice in the rock section by its abundance and perfection of crystal- 
lization is pyroxene, which occurs in dark pea-green crystals. This 
mineral, almost without exception, is twinned along the plane of 
the orthodiagonal. The average of six measurements gave 37° as 
the angle of extinction ; the elements were very variable, possibly 
owing to slight superficial alteration. The crystals are exceedingly 
rich in clear brown glass enclosures, which include vacuoles of gas. 
One crystal showed three such cavities of amorphous matter all tra- 
versed by one large rod-like microlith of pyroxene, the outline of 
which could also be distinguished where enveloped in the interme- 
diate crystalline mass. That portion, where surrounded by the 
glass, had deposited on its surface some microliths of hiotite(?), which 
no doubt separated out from the vitreous matter and crystallized on 
the faces of the prism, though, curiously, none were to be seen on 
the walls of the cavity. 
Magnetite occurs in good-sized unaltered octahedra, sometimes 
enclosed in pyroxene and at others attached in groups of three or 
four on one crystal, themselves supporting hexagonal plates of 
biotite. 
This latter mineral, or black mica, is rather abundant in large 
crystals of the usual characters. 
Sanidine is rather rare in well defined-crystals, often surrounding 
biotite. The most common are irregular, blotchy, or flecked in appear- 
ance when seen by polarized light, as if they had been broken up into 
innumerable fragments and then pressed together, so that unequal 
strains were exerted on each separate morsel. This really seems to 
be the case ; for all are enclosed in a zoned coating of uniform sani- 
dinic matter, which by its contraction in cooling has compressed the 
