531 



two standard fluids. This was, of course, to be done by compensating 

 with one of these fluids the effect produced by the other. "When 

 this is done, it is plain that the length of the columns of the two fluids 

 through which the light passes will be inversely proportional to their 

 rotatory powers. The author did not anticipate any difficulty in effect- 

 ing this compensation ; for, inasmuch as the two species of oil of tur- 

 pentine, chemically identical in their composition, differ very slightly 

 in their refractive and dispersive powers, the law assumed by Biot 

 would lead us to expect that the rotations produced in the planes of 

 polarization of any two homogeneous rays should be in nearly the 

 same ratio in each of these fluids. If this were true, the ratio of the 

 lengths of equivalent columns of these fluids should be nearly the same 

 for all the rays of which white light is composed ; and therefore, when 

 these columns are so proportioned that the intensities of the light in the 

 two images* are the same, the tints will also be the same, so that there 

 should be no difficulty in making the experiment. 



The actual result, however, was wholly different from that which 

 the law of Biot had thus led the author to anticipate. So far from giv- 

 ing a tolerably perfect compensation, the two species of oil of turpen- 

 tine compensate each other more imperfectly than any two substances 

 which the author has examined — in fact, it is scarcely possible to make 

 the experiment at all. When the lengths of the columns of fluid are so 

 adjusted that the intensity of the light shall be the same in the two 

 images, | the tints are found to be wholly different. The left-hand 

 image shows a very brilliant sky-blue, and the right-hand image an 

 equally brilliant rose colour ; the beauty of these colours is very re- 

 markable. There is, therefore, plainly a deficiency of red light in the 

 left-hand image, and a deficiency of blue light in the right-hand 

 image. 



In seeking to interpret this appearance, it is to be remarked that 

 when the light passes through a single fluid whose rotatory power is 

 not very great, left-handed rotation is indicated by the darkening of the 

 left-hand image, and right-handed rotation by a similar effect produced 

 in the right-hand image. Assuming, then, that, in the compound effect 

 produced by transmission through the two columns, the rotation is 

 compensated, or in other words reduced to zero for the mean ray, it is 

 plain that this effect will be left-handed rotation for the red ray, and 

 right-handed rotation for the ray at the other extremity of the spectrum. 

 It is evident, therefore, that the right-handed or Carolina turpentine 

 rotates the plane of polarization of the red ray less, and that of the blue 

 ray more than the left-handed or Bourdeaux turpentine. The Carolina 



* "Proceedings," vol. vii., p. 349. 



f So far at least as this can be effected ; but the truth is, that it is almost impos- 

 sible to compare with any tolerable degree of accuracy the intensities of light whose 

 colours are so different. Even readings taken by the same observer will not agree with 

 each other, and the personal equation will obviously be very large. 



