EXPLANATION OP AN APPARENT NEW POLARITY IN LIGHT." 9 



of the diagram by 30°, the third by 60°, and so on ; and taking the sum of the ordi- 

 nates which are then vertically one below the other. It is evident that the ordinates 

 zero correspond throughout. If we perform the same operation numerically, com- 

 bining the first number of the first column in the table with the fourth number of the 

 second column, the seventh number of the third column, &c., and if we then divide 

 the sum by twelve, we find the following numbers : 



13646, 12829, 11295, 9227, 6875, 4524, 2456, 921, 105, 105, 921, &c.; 

 the greatest number being 13646, or a little greater, and the least being 0. 



It is evident that these numbers denote the formation of most vivid black and 

 bright bands. 



(\e 

 when y is exactly equal to the change of Jc corre- 

 sponding to a change of 2 ^ in RJ is the most favourable for the production of bands ; 



but it will easily be understood that, in consequence of the small extent of the dif- 

 fused image, conspicuous bands may be formed when the change of k corresponding 



to a change of 2 ^ in R is sensibly greater or less than -r. 



The interval between the bands is — r— , and is, therefore, usually small. They will, 



however, be made broader by making h small, that is, by contracting the aperture of 

 the pupil, or by using a telescope with a limited object-glass. The value of R 

 changes through 2 ^r with no greater change in the quality of light than that produced 

 by passing from one part of the spectrum to another part distant (on the retina) by 



—j^, and therefore the retarding plate must be comparatively thick. 



It is evident that these are the bands seen by Sir David Brewster when the spec- 

 trum was viewed in focus. 



The investigation, as regards the explanation of the formation or non-formation of 

 bands under different circumstances, when a thin plate of a transparent medium is 

 placed to cover a portion of the pupil, and the eye is turned to view a spectrum, may 

 now be considered as sufliciently complete, and (I conceive) as perfectly satisfactory. 

 Some change in the expressions would undoubtedly be produced by introducing the 

 consideration of the circular form of the pupil, the inclined position of the transparent 

 plate adopted by Sir David Brewster in some experiments, &c., but none, I appre- 

 hend, which would at all affect the general explanation. 



G. B. Airy. 



Royal Observatory , Greenwich, 



Oct, 23, 1840. 



MDCCCXLI. 



