MONOCLINIC DOUBLE SELENATES OF THE IRON GROUP. 
405 
The extreme readings, for red Li and greenish blue F light, were repeated and con¬ 
firmed several times, and the fact definitely established that the optic axis emerging 
to the left of the normal to the plate was dispersed 7' (within the error of experiment, 
which even with the beautifully small innermost rings and very sharp hyperbolic 
vertices obtained amounts to possibly 2'), while the right-hand optic axis was dispersed 
double this amount, 15'. That the extreme outside reading in each case (the highest 
on the left and the lowest on the right) was for red was clearly apparent, and was 
confirmed by the colours bordering the vertices when seen in white light, the inner 
margins of both being tinted red, and the left one more faintly than the right. 
Hence, it can be accepted with confidence that the second median line was dispersed 
15'_ ,*j f 
to the extent of -———, that is 4'. This is also clearly shown by the mean readings 
Li 
for the position of the second median fine, given in the last column at the foot of the 
table, the difference between these mean readings for Li and F fight being 4'. The 
2' possible error in the readings of the positions for the right and left vertices for any 
one wave-length is eliminated in taking this mean value, and also the mean of the 
two separate sets of observations (crystal behind and crystal in front of the mount 
plate), 3''5 and 4'* 5, is also obviously 4'. Hence, the amount of the dispersion of 
the median fines is undoubtedly 4', with a possible error of only one minute on 
each side. 
In mounting the crystal on the glass plate great care was taken to observe the 
orientation of the crystal, as regards particularly the direction of the little strip-face 
of a (100) replacing th e pp edge, and that of the two faces of r' {201}. The former 
lay to the left and the larger of the two latter to the right, as shown in fig. 3, when 
the 'crystal was behind the glass plate. The edge-face a (100) is parallel to the vertical 
crystal axis c, and this axis emerged 13° 50' to the left of the normal N to c (001), for 
, the plane of the latter face is parallel to the inclined crystal axis a, and the crystal 
axial angle (3 = ac — 76° 10'. Now the reading for the normal N was 214° 18' (see top 
of table), from the observation taken before the plate was wetted with the immersion 
liquid (cedar oil), and the readings for the second median fine (the half-way positions 
between those of the observed left and right optic eye vertices) were shown to be 
217° 37' —41' according to the wave-length (from Li to F). These latter readings 
were also to the left of the normal. Hence, the second median fine is inclined from 
the normal to c (001) towards the left, that is, towards the vertical crystal axis c, to 
an extent which increases regularly from 3° 19' for red Li fight to 3° 23' for greenish 
blue F hydrogen fight. Thus the second median fine is nearer to the vertical axis c 
by 4' for F fight than for C fight. The mean, 3° 21', is the inclination for yellow 
sodium fight of the second median fine to the normal to c (001), and is the extinction 
angle which the author generally determines on the stauroscope, with a couple of 
section-plates parallel to the symmetry plane b {010}. 
The first median fine, the direction of the y axis of the indicatrix, is, of course, 
3 H 2 
