68 H. J. JOHNSTON-LAVIS ON THE GEOLOGY 
The most favourable localities where these deposits may be studied 
are the ridge between the Vallone Grande (Massa di Somma) and 
the Vallone di Pollena, the cupa to the north of the last-mentioned 
valley, and the Valloni Von Buch, St. Patrizio, and Sanseverino. 
Remarks.—The most interesting fact connected with the primary 
material of the pumice is the replacement of the usual pyroxene by 
amphibole. It has been remarked that the latter mineral is most 
common in those rocks that have crystallized under great pressure ; 
whereas the sister species, pyroxene, is to be found most abundantly 
in lavas and other igneous rocks that have consolidated at the sur- 
face or under slight pressure. Although there are many exceptions 
in both cases, yet in general the rule holds good. 
Such facts would lead us to suspect that the amphibole had be- 
gun to crystallize either at great depths, or certainly before the 
actual eruption took place, that is to say when the tension was very 
great. Whether such crystallization was brought about by the 
escape of vapour into thermal water or by conduction, or possibly 
by the low temperature of the original magma, is not easy to 
decide. That the magma was not of high temperature or rich in 
watery matter seems evident from its compact nature in proportion 
to its crystalline structure. Nevertheless the eruption seems to have 
been one of the most violent, if we judge by the amount and charac- 
ter of the ejected materials. It must, however, be admitted that 
all these deductions are of a very speculative character, and there- 
fore must be valued accordingly. 
The peculiarly uniform distribution of this bed wherever it is ex- 
posed in section, and its great thickness, show us the large mass of 
material it represents, spread as it is over a considerable area. 
When we justly appreciate the comparatively small amount of new 
matter ejected, the components of the cone and subjacent rocks con- 
tributing by far the greatest share of the ejectamenta, it becomes 
quite evident that this eruption must have cleared out and consi- 
derably enlarged the preexisting crater. 
We may conclude with good reason that the apex of the new crater 
must have extended downwards through the Tertiary tufas, marls, 
sands, &c., and also to some extent into the subjacent limestone. 
The first rending asunder of the superincumbent obstacles with ex- 
plosive violence was accompanied by the ejection of part of the magma 
itself. Very soon the terrific effort exerted by the elastic vapour was 
partly exhausted, so that a smaller quantity of fluid rock was hurled 
forth, until at last the escape of the gaseous matters was no longer 
able to carry up much of it, although there seems to have been a 
final puff of pumice. The effect may well be compared to the open- 
ing of a soda-water or champagne bottle. On removal of the cork 
the sudden formation of gaseous carbonic acid takes place with 
such rapidity that in its hasty escape it carries much of the fluid 
with it. When, however, the pressure has been reduced consider- 
ably the gas escapes more quietly. 
It is just the latter stage of the above example that is worth con- 
sidering. The vapour rising from great depths carries up no more’ 
