206 AUSTIN F. ROGERS 
(average of 4° 5’ and 3° 15’). This proves the labradorite to 
be about Ab, An,. Fig. 6 is a side elevation of this labradorite. 
The imperfect cleavage parallel to the unit form {11o} or 
{r1o} and j110} in the feldspars has not been sufficiently empha- 
sized. It is present in nearly every specimen of the common 
feldspars which I have ex- 
ANC amined and is a valuable aid 
in orienting cleavages and 
imperfect crystals. Were it 
not for the prismatic cleav- 
age, it would often be impos- 
sible to orient the feldspar 
and hence to determine the 
kind of feldspar. Even when 
not distinct, its intersection 
with (oro) gives the direction 
of the c-axis. 
That this plane of sepa- 
ration is cleavage and not 
Fic. 6.—Labradorite, Ab.An, parting is probable from a 
consideration of the internal 
structure of orthoclase. There are two possible space-lattices 
for monoclinic crystals, the monoclinic parallelepiped and the 
clinorhombic prism. The latter fits orthoclase better, as can be 
seen from Figs. 7a and 7b. According to Bravais, cleavage is 
parallel to the faces of greatest density, that is, those with mole- 
cules most closely packed. From the figures it can be seen that 
cleavage in orthoclase must be parallel to the {oort, {oro}, and 
{110} forms, for the distance between adjacent rows of molecules 
is greatest for these directions. 
The other feldspars evidently have almost the same structure 
as orthoclase. The separation parallel to the unit forms }110} 
and j11o} is cleavage rather than parting, though there may also 
be parting in the same direction in addition. Thus in the albite 
from Auburn, Me., and in oligoclase from New York City, the 
surfaces parallel to (110) and (110) are coated with minute scales 
of sericite and the surfaces are rather dull. 
ob 
ase 
