frames. The polarisation apparatus is, however, so extremely delicate, that unless all the wires are strictly parallel, and the gratings exactly crossed, there is always a resolved component of radiation which acts on the sensitive receiver. It is very difficult to secure the exact crossing of the gratings. I have found it to be a better plan to take two thick square plates of copper of the same size, and, placing one over the other, cut series of slits (which stop short of the edges) parallel to one of the edges. One of these square pieces serves as a polariser, and the other as an analyser. When the two square pieces are adjusted, face to face, with coincident edges, the gratings must either be parallel or exactly crossed. Such accurate adjustments make it possible to carry out some of the most delicate experiments.
The radiator-tube, with the lens and the attached polariser, is capable of rotation. The emergent beam may thus be polarised in a vertical or a horizontal plane. The analyser fitted on to the receiver may also be rotated. The gratings may thus be adjusted in two positions.
- (1) Parallel position.
- (2) Crossed position.
In the first position the radiation transmitted through both the gratings, falls on the sensitive surface, and the galvanometer responds. The field is then said to be bright. In the second position the radiation is extinguished by the crossed gratings, the galvanometer remains unaffected, and the field is said to be dark. But the field is restored on interposition of a double-refracting substance in certain positions between the crossed gratings, and the galvanometer-spot sweeps across the scale.
