438 DR. BUNDJIRO KOTO ON SOME JAPANESE ROCKS. 
SiO, eae ae 53°26 
ATO. 9... BESS 4-01 
Pel. eu. ¢ 3°42 
Pee a ee 14-07 
Min Qt: Peis. 2 trace 
CaOUsEORE Lat ot 10°15 
INO! Crier, Set, 14°65 
99-56 
From the above, we get the following numbers :— 
Sima epee 24-85 
YAS ees tak eee Be 
d Te NaS EN ef 2°39 
Fe 10°94 
Cain Wi Tee f25 
Ae ere a Coke 8°79 
from which it folows:— 
R:Si=l1: 1:19 
R:R=1: 12-40 
R+= AT ale 
Fe: Ca: Mg=1°02: 1: 2°03 
By the Thoulet solution, the augite and magnetite could not be 
separated from each other, the specific gravity of both being higher 
than that of the solution. The larger grains and crystals of 
magnetite were separated from the mixture by the magnet; but 
those which were enveloped in the augites and small granules 
attached to them could not be separated. The high amount of 
sesquioxide of iron in the analytical result should perhaps be 
attributed to this cause. 
From a chemical point of view, this augite has the closest relation 
to diallage. 
Magnetite, of primary as well as secondary origin, occurs in 
large quantity and constitutes one of the essential ingredients of the 
rock. It is found either in grains or in octahedra, and the augite 
is never free from its presence. 
Magnetite is large in its dimensions in proportion as the other 
constituents are also large, but is inversely proportional in the 
perfection of its crystallographic form; that is to say, the coarser 
the other constituents the larger is the size of the magnetite, but 
the more imperfect it is in its crystallographic form. 
Secondary magnetite (in“ contradistinction to the primary) origi- 
nates from the decomposition of augite together with quartz and 
calespar, and this magnetite, as well as that of primary origin, 
passes by decomposition into limonite. 
Quartz occurs in the quartz-bearing augite-andesite in the 
form of grains and never in well-developed crystals. Glass-inclu- 
