52 Dr. Brewster’s experiments on the 
transmitting polarised light through the oil of mace ; but as 
this is not the case,* it necessarily follows that there are two 
halos, or nebulous images, the one lying exactly above the 
other, and having every alternate sector polarised in an op- 
posite manner, while each sector in the one image has an 
opposite polarisation to the corresponding sector in the other 
image. An idea of this curious property may be formed from 
fig. 4, in which we have shown the two halos at a distance, 
and distinguished the opposite kinds of polarisation by the 
signs + and — . 
Those parts of oil of mace which depolarise a portion of 
the bright image, while they form the luminous sectors, have 
therefore the faculty of formin gfour images, two bright and 
two nebulous, possessing the characters which have already 
been described. 
The two adjacent images which are formed by some por- 
tions of the plate of oil of mace, are obviously produced by 
the sectors 1, 7, and 2, 5, fig. 1 being condensed on each side 
of m, and when three images are depolarised, the third image 
is a portion of the bright image restored at m, the place of 
its evanescence. 
If we knew in what way the halo itself is formed, there 
would probably be no difficulty in explaining these remark- 
able phenomena. The diameter of the halo is too small to 
allow us to suppose that the polarisation of the sectors can be 
effected by oblique reflection or refraction, and though it is 
extremely probable that light is partially polarised by inflexion 
• In some instances, when we examine the halo formed by polarised light without 
applying the calcareous spar, the two lectors in the plane of polarisation are less lumi- 
nous than the rest. 
