394 



On the Refraction of Plane Polarised Light. [Dec. 14 r 



tically polarised. The defect is most marked with the object-glass of 

 the collimator. If it be looked at between crossed Nicols in a pencil 

 of parallel rays the field of view is bright, and is traversed by two 

 brushes, hyperbolic in form, which for two positions of the lens 

 become two straight lines, cutting each other at right angles. 



If, again, plane polarised light be allowed to pass through the lens, 

 and only the central part of the lens examined, while it is turned round 

 its own axis I find that for four positions of the lens at right angles to 

 each other the emergent beam is quenched by a Nicol placed with its 

 principal plane at right angles to that of the polarising Nicol, the 

 emergent light is plane polarised ; but for all other positions of the 

 lens the light is not quenched, but reduced to a minimum ; the emer- 

 gent beam is elliptically polarised, and the principal plane of the 

 analysing Nicol is then (according to Eresnel's supposition) parallel 

 to the minor axis of the elliptic vibration. Moreover, the position 

 of this minor axis, as the lens is rotated, changes by about 25'. 



Now, in the experiments described in my paper, it was supposed 

 that plane polarised light, polarised in a known plane, fell on a 

 certain prism of Iceland spar. 



These recent experiments have shown me that the light was not 

 plane polarised, and that even if we suppose the vibration along* 

 the minor axis of the ellipse to be so small that it may be neglected 

 in our theoretical calculations, still the vibration along the major 

 axis will differ considerably in direction from that of the plane 

 polarised beam incident on the lens, while the angle between this 

 major axis and the direction of vibration of the original beam will 

 depend on the position of certain lines in the lens relatively to the 

 plane of polarisation of the incident beam, and will vary as this 

 position changes. 



Moreover, the errors produced for two positions of this plane of 

 polarisation differing by 90° will be the same in amount but opposite in 

 sign, and this was the case in my experiments. It is certain that this 

 defect must modify seriously the results of my experiments, it may 

 possibly reduce to nothing the differences recorded in my paper be- 

 tween the electro- magnetic theory of light and experiment. 



I have found, moreover, that the defect is a common one with 

 lenses. Most of those I have tested since I observed it, some ten or 

 twelve in number, show it to some extent. In none, however, is it so 

 marked as in the one I have had the misfortune to use continuously 

 during the past two and a half years in experiments on polarised 

 light, all of which are affected by it. I am now endeavouring to 

 procure a lens suitable for my purpose free from the defect, with the 

 intention of repeating the experiments, and hope at some future time 

 to lay my results before the Royal Society. 



