242 The American Geologist. October, 1895 
of the two kinds of iron ores which are afforded by the region, the mer- 
chantable magnetites and the titaniferous. The former are in gneisses; 
the hitter in the gabbros and anorthosites of the Norian, which are be- 
lieved to be intruded throiigh the gneisses. A list of localities of the 
titaniferous ores was given and the distinction was made between the 
smaller bodies which are, so far as can be seen, basic developments of 
gabbro, and the enormous ore bodies at the old Adirondack Iron Works- 
in the heart of the mountains. These latter are in massive anorthosite, 
which is almost entu'ely formed of large, blue-black crystals of labra- 
dorite. The largest ore body, which is the one crossing Lake Sanford, 
contains numerous mcluded labradorite crystals, each of which is sur- 
rounded by a reaction rim 5-10 mm. across. It is further noted that the 
wall rocks show no signs of the widespread crushing that is exhibited 
in the general "mortar-structure" of the Adirondack and Canadian an- 
orthosites, but are plutonic rocks, free from evidences of dynamic met- 
amorphism. The argument is then made that the ores are segregations 
from an igneous magma formed during the process of cooling and crys- . 
tallization. In conclusion the speaker gave some notes on recent at- 
tempts' to utilize the titaniferous ores that bid fair to be successful. 
In discvission. Prof. Van Hise mentioned the similar bodies of titanif- 
erovis ores m the gabbros of lake Superior, adding, however, that there 
had been some infiltration of iron oxide since their formation. 
The Decomposition of Rocks in Brazil. J. C. Branner, Stanford 
University, Cal. The deep decay of rocks in Brazil is notorious, thovigh 
but few observations have been published on the subject. The present 
paper embraces the results of the author's observations made during 
the eight years he lived and traveled in that country, together with the 
statements of other geologists. The evidences of deep rock decay are 
found in railway cuts and tunnels, excavations in hills in Rio de Janeiro 
for buildings and for a reservoir-, deep mines in the gold region of Minas 
Geraes, in enormous gullies of recent origin, and in numerous landslides. 
Many instances of decomposition to a depth of more than 100 feet are 
known, and in some of the old gold mines the rocks are soft to a depth 
of more than ,300 feet. This decomposition seems to be widespread, 
though not universal, in Brazil. 
Exfoliation is a common feature, and is not confined to boulders: 
massive rock hills and mountains exfoliate in the same way as the boul- 
ders of decomijosition. Some of the peculiarities of the topography in 
the granite and gneiss regions are due to this method of decomposition. 
Talus slopes ai"e very rare. Massive rocks sometimes weather into flu- 
ted surfaces, having steep-sided trenches and ravines that run straight 
down the rock faces. 
The chief mechanical agency promoting rock decay is change of tem- 
peratvire within a range of about lOO'* Fahr. But little direct work is 
done by this agency, and its chief importance lies in the fact that it 
opens crevices that admit the chemical agencies of decay, gases, water 
and acids, which destroy the rock rapidly. Color is believed to be of 
some influence in this coimection, as the black crvstalline rocks are sel- 
