22 J. M. Blake on Hortonolite. 
red rays was 83° 41’, and 80° 29’ for the blue, measured in oil. 
The bisectrix in this case being negative, and normal to g’. 
For comparison, a measurement was made of olivine occur- 
ring in lava from the Sandwich Islands, and the oe aa of 
the axes about the normal to g' was fi 
nace product 68°. Both were measured in olive oil. Redifting 
the measurement on the chrysolite, from Torre del Greco, to the 
same bisectrix, but without making a correction for the “refrac- 
tion of the oil, and of the mineral which would be necessary 
to make the results compare accurately with the others meas- 
ured in oil, we have 94°. 
It so. far appears that the optical axes in this group 
lie in a plane parallel to h', but vary in separation with the 
chemical composition, according to some law. It seems gene- 
rally preferable to make sections parallel to g', for the pur- 
pose of comparing species ; since this is usually a cleavage 
plane, and there is then a better chance of success in cutting 
some varieties which are brittle, and the plane p seldom rp 
while g' is often well defined, and serves as guide. It is 
be understood that reference is made to such sections as can 
most readily cut for the a aa - comparing varieties of this 
mineral, as is done with the mic 
The cleavage which occurs parallel to h',is perhaps most 
clearly defined. There are two cleavages about the plane g1, and 
possibly one not well defined parallel to this plane. They lie 
in the prismatic zone, g, e', etc. This was first discovered 
in the following way: a fragment showing several cleavage 
surfaces, and at least one plane of the crystal, was measured 
angle of the normals of the two cleavages. Hence the cleav- 
ages are symmetrically situated about the plane g', and in the 
zone, g' es ef, etc. ; since another cleavage was found on this _ 
a Legare’ ular to the intersections of the two others, 
giving as an approximate measurement an inclination of 92° 20! 
on one of hy and this cleavage had also a different appear- 
ance to the eye, and was therefore considered to be the cleav- 
age parallel to the basal plane h'. No stale from a 
cleavage was detected on this specimen parallel to g'. The 
three cleavages actually found gave bright reflections in a good 
light, but admitted of no very exact measurement ; though 
sufficiently exact to show that those on the prism are ‘not par- 
allel to any observed plane, except it be the rarely neers 
planes es ore}. The inclination of ef to g', is given as 2 
—— | 
vey — 
