Gabhroid Rocks of Minnesota. — VVi7ichclL 179 
marked by the appearance of diallagic parting planes which 
always appear first on the outside of the crystal and develop 
inward. In their characteristic development the parting lines 
are very fine and close together, but they seem to occur rarely 
as a series of coarse fracture lines similar to the coarse pyro- 
xenic cleavages, being merely more irregular. The augite 
sometimes passes through the diallage stage before altering 
to other minerals, but about as frequently alters directly to 
chlorites, micas, hornblendes, etc. 
Chloritization of the pyroxene is very common. The pale 
green chlorite fibers usually appear first about the edges of 
the augite, but rapidly propagate themselves along the cleav- 
age and fracture lines. The pyroxene may thus be wholly 
transformed to a mass of fibrous penninite or clinochlore, or, 
more rarely delessite*. 
Uralitization is not uncommon, and the resultant horn- 
blende is sometimes brown, and sometimes green. It may be 
either fibrous, or distinctly more compact, and lamellar. It is 
always pleochroic. 
It is uncommon to find the pyroxene altering directly to 
biotite, without the evident cooperation of the feldspar. The 
formation of the mica is therefore usually in the nature of a 
"reaction" between two minerals, and one of these two is al- 
ways labradorite, while the other is oftenest magnetite, but 
may also be augite, or, more rarely, olivine. The biotite oc- 
casionally occurs, however, in fibers along the fractures of 
pyroxene while the surrounding feldspar is wholly unaltered;, 
in such cases it is evident that the mica is the result of the 
alteration of the augite alone. 
Finally, all these various methods of alteration of the 
pyroxene are often accompanied by the separation of grains 
of iron oxide, indicating that the augite was originally highly 
ferriferous. These grains are usually magnetite, but also oc- 
cur as hematite. 
The serpentinization of the olivine is an alteration so well 
known that it is only needful to state that the resultant mineral 
is bowlingite much oftener than antigorite. The former fre- 
*This mineral has not been observed by the writer, but is a com- 
rnon mineral in the lake Superior region, and is noted as a decompo- 
sition product of augite by Bayley. See Bayley: Op. cit. Jour. Geo!. 
I- 1893, p. 704. 
