OF MONTE SOMMA AND VESUVIUS. 61 
to form one of its most important constituents. The crystals are 
generally well formed, occurring singly or in groups. As a rule the 
mass is pretty free from impurities, but occasionally trains of dirty 
glass cavities are met with, or sometimes there is an intimate mix- 
ture of this felspar with pyroxene, which seems to be an advanced 
stage of what one observes in the earlier deposits, and was mentioned 
when describing the intermediate pumice between Periods 1 and 2 
of Puase III. Some crystals of felspar have the triclinic character ; 
they are numerically inferior to the monoclinic. 
The microliths of pyroxene range from small crystals downwards. 
Many of the larger of these form stellate bunches, which sometimes 
project in a fan-like manner into the vesicles, where their crystal- 
line form may be seen to perfection. Magnetite microliths are 
large and well developed, being four or five times the size of those 
in the preceding pumiceous scoria. They are much more altered, 
and have, as a consequence, lost their crystalline outlines, and ap- 
pear as irregular rounded masses. They also form a dense black 
wreath around the sanidine sections. 
The decomposed aspect and colour of the lava seem principally 
due to this alteration of magnetite. 
The small microlithic plates of biotite are very perfect, and occa- 
sionally project into the cavities, like the pyroxene. The larger 
elongated microliths are sanidine; although their terminations are 
brush-like, the Carlsbad twinning is very common and distinctly 
developed. The smallest rod-like microliths are probably pyroxene, 
but have no distinctive characters. 
There are in the section one or two circular (?) spaces that darken 
between crossed nicols, although apparently not vesicular cavities, 
and therefore they may be minute crystals of leucite. They are too 
small, of course, to show the optical characters of that mineral. 
The great resemblance of the lava in the Vallone Von Buch to 
that just described, their identity of stratigraphical position, and their 
similar microscopical structure and state of decomposition, incline 
us to regard them as nearly, if not quite, contemporaneous, although 
their wide separation from each other shows they were not the 
same stream. 
Puasst IV., Period 2. 
General Description.—In the Vallone Sanseverino, above the last- 
described lava-flow, is another, differing considerably in structure 
and other characters. It is well seen at the cascade, of which it 
forms the ledge, near the contour-line of 325 metres. 
This upper intrapumiceous lava is of a bluish-grey colour, very 
vesicular and ragged. It is also very tough, and breaks with much 
difficulty. It differs from the lower stream by the larger and fuller 
development of the enclosed crystals. The scales of black mica are 
sometimes two or three millimetres across, and are pretty abundant, 
giving to this rock quite a peculiar appearance. This lava shows 
none of that cleavage parallel to the surfaces seen in the earlier one. 
Microscopical Exanunation.—The general structure of this upper 
