184 FRANCIS DEAS ON THE SPECTRA FORMED BY THE PASSAGE OF 



segment to be replaced by the coloured rings of the next. If we again circu- 

 larly polarise the light by interposing a second circularly polarising plate between 

 the crystal and the analyser, the brushes entirely disappear, and both the black 

 and the coloured rings are continuous throughout, forming perfect circles. 



When the analyser is rotated through 90°, the centre of the system which 

 was formerly black is now coloured, and, at the same time, all the black rings 

 have exchanged places with the coloured rings, the change being effected by a 

 lateral displacement in opposite directions of the two halves of the circle. 



If, for the second circularly polarising film, we substitute a film of a different 

 thickness, the rings assume curiously distorted forms. With one film which I 

 used the rings became ellipses, with another they all united so as to form a 

 circular helix, which appeared to unwind like a screw as the analyser was 

 turned. 



The appearances produced by using different crystals are, of course, similar 

 mutatis mutandis. 



By circularly polarising the light before and after its passage through a 

 crystal of nitre, the brushes are wiped out, and the lemniscates are beautifully 

 seen, unbroken throughout. 



When a crystal of Iceland spar is used, and the Nicol's prisms set with their 

 axes inclined 45° we get eight segments, of which the four light segments look 

 as if they stood out in relief against the dark segments, while the sections of 

 the black rings, especially near the centre of the system, look more like straight 

 lines than circular arcs, and form a system of octagons. 



The effect upon the rings, produced by placing on the stage a film of selenite 

 in the position in which it should give the black bands previously described, 

 is a strange one. Instead of a black band occurring, the coloured ring 

 belonging to that part of the spectrum is seen to split into two. It sends off 

 a branch as it were from its lower part, which shoots across the adjoining 

 black ring, and joins itself with the lower part of the next coloured ring. 

 This last ring then in turn sends off a branch from its middle part, which in 

 like manner unites with a third ring, which in turn does the same to its neigh- 

 bour. All this takes place within the space which should be occupied by the 

 black band. 



The beauty of these last experiments, wonderful as it is, may be still further 

 enhanced by the use of a double image prism as the analyser. The result is 

 analogous to that obtained with the same arrangement in the case of the selenite 

 previously described. We now get not only two spectra but two systems of rings 

 which, by superposing the spectra, may be made to interlace with one another. 

 Wherever a black ring of the one spectrum crosses a black ring of the other, 

 the intersection is of course still black. Where a coloured ring of the one 

 system crosses a black ring of the other, it retains its original colour ; but if a 



