THE VOLCANOES AND ROCKS OF PANTELLERIA 687 
could be identified by me. The hornblende is a very dark chest- 
nut brown, intensely pleochroic, in small prisms and anhedra, 
which show no cleavage. It is probably cossyrite, as suggested 
by Rosenbusch, or possibly kaersutite, which has been found on 
Linosa. The latter is suggested by its pure brown color, without 
the reddish tinge so characteristic of cossyrite, and the nearly or 
quite parallel extinction of the small prisms in some specimens 
from Montagna Grande. The hornblende frequently surrounds 
and is later than some of the groundmass feldspars, while the 
pyroxene belongs to an earlier period of crystallization. 
No magnetite is present in the groundmass of any of my speci- 
mens, but flakes of orange or blood-red hematite occur in some 
cases, which do not seem to be secondary. A little quartz may be 
detected in some cases as an interstitial residuum, but its amount 
is small. 
Hyaline facies——A specimen from the surface of a flow at 
Costa Zichidi, which is somewhat vesicular, is different from the 
others. In thin section are the usual feldspar phenocrysts, with 
rounded ones of red-brown cossyrite, a few of aegirite-augite, but 
none of augite. The groundmass consists largely of a pale yellowish 
glass, containing minute laths of feldspar. This glass is thickly 
crowded with peculiar dendritic growths of a deeper brown color 
which often assume a globular form from 0.2 to 0.5 mm. in diame- 
ter, and again larger irregular areas. Feldspar phenocrysts 
usually form the nucleus of these. They are made up of small 
grains seldom more than o.o1 mm. in diameter, either equant or 
elongated. 
Toward the center of the spherulites the grains are thickly 
crowded, but they separate at the borders, which consist of small, 
irregular, radially divergent tongues and processes. The grains 
are light yellow in color with rather high birefringence, the areas 
showing a faint aggregate double refraction. Their extremely 
small size makes optical determination difficult, but they are 
probably an aegiritic augite. 
Chemical com position.—Analyses were made of a specimen from 
a massive flow near the base of the southwest scarp of Montagna 
Grande, below Rione Miliac, of one from the crater of Monte 
