OPTICAL, CHARACTER OP BIREPRACTING MINERALS. 79 
in all positions for the usual microscopic field of view. Mallard* was appar- 
ently the first to direct attention to the fact that the convex side of this 
dark axial bar always points toward the acute bisectrix. Since Mallard's 
time this method has been used and described by a number of writers, 
including F. Becke in 1904,! the writer in 1905,! and C. Ce*saro in I9o6.|| 
From the descriptions it is apparent that each of these investigators dis- 
covered this method anew and independently of the others. 
Having once determined the convex side of the axial bar in the section 
normal to an optic axis the optical character of the mineral is then most 
readily ascertained by use of the sensitive-tint plate. If the achromatic 
bar be placed in the position of Fig. 49 
with its convex side pointing to the north- 
west, and the arrow of the sensitive-tint 
plate (= C, direction of least ellipsoidal 
axis) be in the same direction, the convex 
side of the curve will show a blue inter- 
ference color if the mineral is optically 
negative; if it be optically positive, the 
blue spot or border will appear on the con- 
cave side of the axial bar. 
This method can always be applied if 
the convexity of the curve is well marked. 
In certain plagioclase feldspars the limit- 
ing case of 2^ = 90 is encountered occa- 
sionally and there the bar is in fact a 
FlG - 49- straight line. If the section be not cut 
precisely normal to the optic axis, the optical character can still be safely 
determined if the curvature of the axial bar is well marked and the optic 
axis is not too near the margin of the field. ^[ 
PLATE PARALLEL TO THE PLANE OF THE OPTIC AXES. 
In uniaxial minerals this plate corresponds to any section parallel with 
the optic axis. 
The interference figure from such a plate is readily recognized by noting 
that on rotating the stage of the microscope the field becomes suddenly 
dark, remains so for an instant, only to become light again on further ro- 
tation of the stage through a few degrees. In the position of darkness the 
greatest and least ellipsoidal axes, ct and C, are parallel with the principal 
sections of the nicols. No distinct cross is seen, as in the interference figures 
of plates normal to the bisectrices. The entire field appears dark, with 
perhaps a weak fringe of light along the outer diagonals of the quadrants. 
If the field be placed in the dark position and turned slightly, faint, dark 
hyperbolas can be seen to open and leave the center of the field, similar to 
*E. Mallard. Traite de crystallographie. Tome II. 229. 1884. 
tRosenbusch-Wiilfing, Microskopische Physiographic I, I, 335. 1904. 
tAmer. Jour. Sci. (4). 20, 292, 1905. 
^ Bull, de 1'Acad. roy. de Belgique (Classe des Sciences). 323-324. 1906. 
lln 1907 the writer endeavored to determine graphically the degree of curvature of the axial bars for 
different positions of the optic axis in the field by the direct application of Fresnel's rule. In the curves 
thus prepared (Atncr. Jour. Sci. (4). 24, 338-340, 1907) the rotation of the crystal plate and intervening 
glass plates and lens on the plane of polarization of the transmitted waves was disregarded as a negligible 
quantity and the curves are inaccurate to that extent. They prove, however, that on sections showing 
an optic axis in the field the optical character can be determined by this method, provided the axial bar ia 
noticeably convex. 
