THE MEDITERRANEAN NATURALIST 
3:38 
for lime or alumina. He attempts to explain this 
by a theoretical principle which he calls that of 
minimum paste, which would not have been re- 
quisite had the physical conditions been taken into 
account. Again, this theory in its incomplete 
form is proved insufficient by the joint author, 
Prof. E. Hull, (1) in the same memoir, although it 
was undoubtedly a great step in the direction of 
an important principle. 
M. Bourgeois (2 ) accounts for the crystals of 
pyroxene in leucite to be the crystallization of the 
glass cavities. This is obviously not the case, for 
the following reasons: — In the leucites of Rocca- 
monatia and Vesuvius the crystals of pyroxene 
entirely traverse, project their ends on each side, 
whilst the leucite material is accurately moulded 
on the crystal facets of the pyroxene, which form 
leucite could not give to a glass space. Besides, 
many pyroxene crystals bear no relation whatever, 
either in size or position, to the remaining cavities, 
which themselves clo not show such crystallization. 
Their crystals are often imbedded in the leucite 
mass, and project into a glass cavity, the latter 
portion being no thicker than the former, which 
was entirely enveloped in the leucite mass. Where 
much growth of crystals in glass cavities take 
place, that portion surrounded by the vitreous 
paste of the glass cavity should have increased in 
size. That the .artificial conditions employed in 
the laboratory fairly represents the natural ones 
in the production of leucite there exists little 
doubt; the variations in temperature were just 
such as we meet with in the formation of that 
mineral at Vesuvius. Besides, the two minerals 
were identical in crystallographic characters, both 
externally and internally, as seen by polarized 
light, and also the great resemblance as exhibited 
in the strata of glass cavities. 
That leucite may separate or any rate increase 
in size, after expulsion of lava, seems to be 
demonstrated bp the observation of Scacchi, (3) 
that the scoria of the lava of 1855 did not contain 
large crystals, and that in the lava the distribution 
(1 ) Op. cit.,p. lJfl. 
(2) Encycl. Chim.,vol. ii . , Metalloids, lev Ap- 
pendic v. Re prod. Artif. des .Roches, p. 212. 
(3) Guarini, Palmieri, Scacchi. Mem. Sid. 
Incend, Vcsuv ., 1S55, p. 132. 
of them was irregular, which seems to show that 
remit at least increased their size. 
In describing leucite I have considerably erred 
from the direct road, led on by the train of argu- 
ment, based principally on the physical and- 
chemical properties of this interesting mineral. 
JBiotitc, though commonly met with in volcanic 
rocks, could not be obtained as a distinct form by- 
Messrs. Fouque and M. Levy-. In lavas we gen- 
erally 7 meet with this mineral in large, well- 
formed crystals, as also in pumices. In some basic 
pumices of Monte Somma ( Pha<> III. ) very beau- 
tiful hexagonal microlithic plates, and small 
crystals may be seen scattered throughout the 
magma, and often enclose crystal s'of orthoelase. In 
the more highly crystalline pumices and lavas this 
mineral occurs generally 7 as well-formed crystals. 
Although it is not very uniform in its occurrence, 
I am disposed to regard it rather as pre-eruptive 
in formation, or at any rate, in part. 
Magnetite is another mineral that cannot be 
obtained by simple fusion, but requires solution in 
a fused medium, from which it separates during 
cooling within a great range of temperature. (1) 
provided the formation of ether minerals renders 
the magma supersaturated, from time to time, 
with this oxide, so that various crops of crystals 
may result, forming so many pen ods o f >».<■, /illat- 
ion. This is the only 7 way we can explain its for- 
mation as with quartz, leucite, Ac. Scheerer 
pointed out long ago the grauite-forming minerals 
separated inversely to their fusion-points. 
Pyroe en, , as well known, is a common product 
in furnace slags, and is easily obtained by simple 
fusion of its elements with a very short remit. 
Messrs. Fouque and M. Levy found it to be produ- 
ced in a microlithic condition after a few moments'' 
remit, and prolonging this a little, hue oy stals, 
such as are met with in volcanic rock, were obtain- 
ed. Such a fact convinces us of the extreme 
rapidity with which basic pumices, at any rate, 
must have passed from the fluid to the solid con- 
dition, as in many of the Italian basic volcanoes 
the first products of some of their explosive 
eruptions were practically without even microliths 
of pyroxene, striking examples of which are to be 
met with in the deposits of Phase , III., period 1, 
(1) Bull. Soc. Geol. 2c serie,tom . iv. page .{75, 
