i8o The A?nerica?i Geologist. September, i90u 
quently orients itself upon the olivine; in such cases a single 
crystal of bowlingite is formed from a single crystal of olivine; 
at other times the alteration mineral is finely fibrous, and has 
all possible orientations, and zeolitic extinction. It is not 
rare to find a crystal of olivine altering to both the fibrous and 
the compact types. Bowlingite varies in color from green 
to brown according to the condition of oxidation of the iron; 
and is nearly always pleochroic. The bowlingite is only rarely 
seen to form away from the olivine and then only at very short 
distances away, but antigorite occurs filling all the fractures of 
the labradorite all the way from one olivine grain to another. 
If any cavity occur, it is also often filled with antigorite. (See 
Plate VIII, Fig. i.) Chrysotile and xylotile perhaps occur, 
but they have not been identified by the writer. 
The only other alteration of the olivine which has been 
observed is to biotite. This is comparatively quite rare. The 
biotite often starts as a decomposition product of the pyroxene, 
and extends into the olivine; in other cases it is a "reaction" 
product to which both the feldspar and olivine contribute some 
elements; finally it seems to occur occasionally from the direct 
alteration of the olivine alone. 
Magnetite is often seen surrounded by a zone of biotite, 
and this is the commonest source of the brown mica of the 
rock. The biotite is evidently in such cases the result of the 
simultaneous alteration of magnetite and feldspar, as it never 
extends away from either. 
Leucoxene has never been observed, though the magnetite, 
at least in certain diabases is known to be titaniferous. In 
highly altered areas small grains of quartz occasionally occur. 
They are doubtless derived from the labradorite, though their 
direct formation from the feldspar has not been observed. 
The chemical compositioii of the normal gabbro varies 
within rather wide limits; nevertheless analyses from widely 
separated localities are quite comparable, the differences being 
readily explained by the comparative abundance of the es- 
sential mineral constituents. The specific gravity generally 
ranges between 2.70 and 3.00. A sample from Birch lake 
which was analyzed is not quite typical, as it shows a marked 
tendency toward the ophitic texture. The column I. shows 
the results obtained, while II. is the average of two analyses 
