34 THE JOURNAL OF GEOLOGY. 
It must further be remembered that the green hornblende 
was itself formed and existent in a magma at a temperature 
undoubtedly much higher than that of the simple Bunsen burner 
employed by Belowsky; but whether in the absence of oxygen 
or not we cannot perhaps say, though in view of the well established 
presence of water in most lavas and its probable dissociation at 
the high temperature of the liquid lava in the conduit oxygen 
may reasonably be supposed to have been present. At any rate, 
in the case of two lavas erupted under such similar conditions as 
those of Stavro and Chondos there is no ground for assuming 
that the latter was at a higher temperature than the former,’ or 
that the one was oxygen-free and the other not. Again, it is 
difficult to explain by Belowsky’s theory the presence in the same 
specimen of crystals of both varieties, which has been frequently 
observed, while if we grant to each variety a separate existence, 
this fact is easily explicable on the basis of the frequent presence 
of streaks (Schhere) in molten magmas, and their consequent 
want of homogenity in chemical composition. 
The plagioclases of the andesites show the same difference 
between the phenocrysts and the groundmass crystals as in the 
preceding rocks. The phenocrysts are usually quite clear, and 
while glass inclusions, often with bubbles, are common and fre- 
quently zonally arranged, they seldom show such well developed 
net forms as have been described above. The phenocrysts show 
much zonal structure and many twinning lamelle, which, giving 
extinction angles of 33° and 34°, show that they are of anorthite, 
while the smaller lathes, giving angles of 22° to 24°, are of 
bytownite. 
Phenocrysts of biotite are not infrequent, and are in most cases 
entirely altered to an opacitic aggregate similar to the hornblende. 
Magnetite is abundant, but calls for no special remark. 
Hypersthene-Andesites. 
abundant presence of orthorhombic pyroxene, hypersthene, which 
These rocks are characterized by the 
tAs the Chondos rock seems to be more basic than that of Stavro, it is even prob- 
able that its temperature on eruption was less than that of the latter; this would also 
hold good for the brown hornblende-bearing hypersthene-andesites and the green 
hornblende bearing dacites. 
