66 H. J. JOHNSTON-LAVIS ON THE GEOLOGY 
one side of the cavity, though of course the chemical action in one 
and the vital action in the other are quite different. So common 
is this enclosure of calcareous fragments that it may be regarded as 
one of the less important characters of this rock. 
Usually the pieces of pumice have been broken at the time of 
their deposition, but it is not uncommon to find some entirely 
covered with their thin but dense crust. The fragments usually 
vary from the size of a hazel-nut to that of a walnut, and more rarely 
they are as large as a man’s fist. 
The usual dirty-brown colour of the matrix is broken occasionally, 
besides the white calcareous enclosures already described, by white 
glassy sanidine crystals and black amphibole. 
Microscopical Examination—The most abundant of the large 
crystals are those of a pyroxene-like mineral. Its sections appear 
black unless very thin, when they transmit a dirty neutral-green 
tint. Rotated above the polarizer their polychroic character is most 
marked, often ranging from a light greenish yellow to a dark dirty 
bluish-green colour. The cleavage-planes, very abundant, combined 
with the outlines of the crystals, point to this mineral as being 
amphibole. This is fully confirmed by the angle of extinction ; the 
maximum I measured was 183°, but it ranges from that down 
poslOss: 
The next most abundant mineral is sanidine, which is scattered 
through the matrix in small and occasionally large crystals, and is 
sometimes attached to and may form wreaths around the amphibole. 
Magnetite is pretty abundant, and occurs in two if not three 
forms; large crystals, well formed, some loose in the matrix, but 
more often forming a nucleus to, or at least enveloped in, the amphi- 
bole. In some specimens these are decomposed, and convert the sur- 
rounding rock into an irregular mass of dirty brown colour, in which, 
however, its true constituents can be distinguished. In one specimen 
examined the magnetite occurs as very small but well-formed octa- 
hedra, many of which are more or less peroxidized. The third 
variety will be discussed with the microliths. By employing a large 
quantity of the rock, and treating it by Fouqué’s method with hydro- 
fluoric acid, a number of crystals can be obtained of amphibole, 
with a few of pyroxene and olivine (?). The two latter, from their 
rarity and imperfect form, may be assigned as erratic in origin, 
having been taken up and enveloped with other foreign fragments. 
Scattered through the matrix rather sparingly are a number of 
well-formed rod-like microliths, a few appearing to belong to 
sanidine (?), amphibole (?), and pyroxene(?). Their characters are 
very poorly marked. Next we have great quantities of minute 
microliths, or perhaps more properly trichites, arranged in sheaves, 
or so as to appear like a long-fibred velvet brush. Hach individual 
component one appears as a dark line, often curved, and does not 
transmit light between crossed nicols, unless when in a very dense 
bunch. Then we may observe a feeble greyish patch, which might 
be compared to the effect produced by a half-dry white-paint brush 
smeared over a black board. The third microlithic substance, differ- 
