702 HENRY S. WASHINGTON 
grains of feldspar and quartz, though colorless glass is present in 
small amount in some sections. 
Pantellerite pumice —With this type of pantellerite may be 
briefly described a pumice which covers a considerable area near 
the center of the island, about the summit of Montagna Grande and 
Cuddia Mida, between Monte Gibelé and Serra Ghirlanda, and in 
isolated patches elsewhere. It is whitish or cream colored, becom- 
ing a brownish yellow on wetting; highly vesicular, floating easily 
on water. Equant phenocrysts of alkali feldspar (2-5 mm.) are 
rather abundant, but none of mafic minerals. A flow texture is 
usually well developed. 
In thin section this pumice is seen to be perhyaline. A few 
of the feldspar phenocrysts, and still fewer and smaller ones of 
aegirite-augite are present. The glass base is clear and colorless, 
carrying only extremely small amounts of “‘dust”’ and very rare 
microlites of feldspar and pyroxene: No hornblende is present. 
Chemical com position.—Two analyses of this type of pantellerite, 
and one of the pumice were made. With them are given two by 
Foerstner and one by Abich of pumice. - 
These are all consistently higher in silica and ferric oxide, 
and lower in alumina, lime, and magnesia, than the trachytes 
described previously. The only difference of note between my 
two analyses of the type is the alumina, the higher figure in B 
being partly compensated for by the slightly higher iron oxides 
of A. Otherwise they are almost identical. It will be observed 
that TiO, is high for such silicic rocks. Attention must be called 
here to the relation of the iron oxides, the percentage of Fe.O, 
being much higher than that of FeO in both. The analysis of 
the pumice is closely like the others, though lower in silica, and 
the proportion of ferric oxide is greater. It is chiefly on this 
account and because of the colorless glass that it is placed with 
the Santelmal type. The analyses of Foerstner are unsatisfactory 
in their incompleteness, and also seem to suffer from the usual 
systematic errors. It will, however, be observed that his iron 
oxides, though higher than mine, show much more Fe,O, than 
FeO. While the analysis of Abich is, of course, unsatisfactory 
according to modern standards, yet, considering its date, it is very 
creditable, the alkalies alone being improbable. 
