EXPLANATION OF PLATE XXI. 
In the accompanying Plate (XXI.), fig. 1 shows the effect produced on the violet line of cadmium 4678. 
At the top, a, we have the line photographed with the magnet unexcited, that is, in the free field. 
Underneath this, at 6, the same line is photographed with the magnet excited, but the field is not strong 
enough to resolve it into its three constituents. It accordingly appears to be merely broadened by the 
magnetic field. A Nicol’s prism was then introduced into the path of the light and the line photographed 
in the same magnetic field, with the result shown at ec, where the middle is seen to be removed from the 
affected line, so that it appears as a doublet. The nicol was then turned through a right angle and the 
line again photographed in the same field. The result is shown at d, which proves that the sides of the 
broadened line have been cut off, while the middle has been allowed to pass. This agrees with the 
supposition that the magnetic field resolves the line into a triplet, but it does not prove it. 
The further resolution necessary to prove this point is shown in figs. 2 and 3. In fig. 2 a photo- 
graph of the zinc lines 4811, 4722, and 4680 is shown, and it will be observed that 4680 shows as a pure 
triplet, while the others do not. Fig. 8 is a photograph in a still stronger field taken from a spark passing 
between two electrodes, one of cadminm and one of zinc, so that the lines of cadmium and zinc are obtained 
simultaneously under precisely the same circumstances. It will be seen that the lines most affected are 
4678 of cadmium and 4680 of zinc, and these both show as pure triplets, while the lines 4722 and 4800 
show as quartets. 
