diameter. No change was made in the bundle A which was kept as a test one. The bundles B and C were then twisted, B in a right-handed direction and C in a left-handed direction.
The interposition of the untwisted bundle A between the crossed polariser and analyser did not produce any effect, but strong action was produced in the receiver when the bundles, twisted to the right or to the left, were so interposed. It thus appeared as if the twisted structures produced an optical twist of the plane of polarisation.
Further experiments to be described below may be of some interest in connection with the optical rotation produced by liquids. Here two different classes of phenomena may be distinguished:—
The difficulties in the way of explaining rotation produced by liquids are summarised in the following extract:—
"It is, perhaps, not surprising that crystalline substances should, on account of some special molecular arrangement, possess rotatory power, and affect the propagation of light within the mass in a manner depend-
