Louw ik DLL OW TAN AL OLIN GTA OM OGNE 
shown by chemical analysis to be low in silica or else described 
as anorthite lavas, while the latest outflows of Giorgio Kaimeni 
belong to the most acid pyroxene andesites.*? But here again we 
have no trace of either basalt or rhyolite. As will be shown 
farther on, the rocks of Nisyros are probably similar to those 
described here, but scarcely anything is known of them. 
Taking all the above facts into consideration, viz., the per- 
manently extinct condition of Kos and Milos, and the almost as_ 
certainly extinct condition of our region, the total lack of basalts, 
almost total lack of rhyolites, and the overwhelming preponder- 
ance of intermediate andesites shading into dacites or very acid 
pyroxene-andesites,3 and the not very basic character of the 
segregations, we may conclude that, while Iddings’ law holds 
good for this region, it does so only partially ; in that as the . 
eruptions proceeded the magma became more and more differen- 
tiated, but that this process of differentiation never reached (as 
far as we can prove positively from the evidence of erupted 
materials), and probably never will reach, the extreme stages of 
basalt and rhyolite, the limits reached here being in one direction 
andesite with about 54 per cent. of SiO, and in the other dacites 
or andesites with 68—6g per cent. SiO,,. 
THE REGION AS A PETROGRAPHICAL PROVINCE. 
The scope of this paper as well as lack of space forbid my 
entering here into a complete discussion of the characters of the 
whole /Egina-Nisyros fracture line, but some remarks may be 
made on its general characters regarded in the light of a “ petro- 
graphical province.” As has just been pointed out the larger 
part of the eruptive rocks along this fracture line are pyroxene- 
andesites, with smaller quantities of hornblende-andesites, which, 
as the magma becomes more and more differentiated, run into 
highly acid andesites or else dacites, with perhaps some rhyolite. 
t FOUQUE: Santorin, pp. 334, 341. Cf. RoTH: Geol. II., 362. 
?Cf. RoTH: Beitr. Petr. 1869, p. cxxiv. 
3It may be here remarked that as far as the published accounts and my own 
observations permit me to judge the eruptions near Smyrna and Pergamon were like- 
wise confined entirely to andesites, and they are now certainly extinct. 
