184 ^Astronomical and Nautical Collections, 



are polarised in orthogonal directions; and consequently 

 the groups of fringes on the right and left cannot exist, un- 

 less we restore their mutual influence by reducing them to a 

 common plane of polarisation : and this is done by the inter- 

 position of the pile of glass plates, or of the rhomboid. The 

 fringes thus obtained are so much the more marked, as the 

 two contrary pencils which produce them are more equal in 

 brightness ; and it is for this reason that they are best dis- 

 tinguished when the principal section of the rhomboid makes 

 an angle of 45° with the axis of the plate. When this prin- 

 cipal section is either parallel or perpendicular to that of the 

 plate, the rays refracted ordinarily by the plate, pass intirely 

 into one of the images, instead of being divided between the 

 two ; and all the extraordinary rays pass into the other 

 image : so that the rays of the separate descriptions do not 

 interfere with each other ; and the additional groups disap- 

 pear : each image presenting only those fringes which are 

 derived from rays of the same kind, that is, those which form 

 the middle group. 



These two groups of additional fringes exhibited by po- 

 larised light, in the first position of the rhomboid, furnish one 

 of the most accurate modes of measuring double refraction, 

 and of studying its laws. In fact, their eccentric position 

 depends on the difference of the paths of the ordinary and 

 extraordinary rays which have passed through the plate, and 

 we may judge of the number of undulations that the extra- 

 ordinary rays of the pencil have been left behind the ordi- 

 nary rays, by the number of breadths of fringes compre- 

 hended between the middle of the right hand group and that 

 of the middle group : and this difference in the paths is still 

 better measured by the interval comprehended between the 

 middle points of the lateral groups, which is twice their 

 distance from the middle of the central groups. It is most 

 convenient to employ white light for these experiments : first, 

 because it is the brightest, and secondly, because it renders 

 the middle point of each group more easily distinguishable. 

 It is true, that we only measure in this manner the double re- 

 fraction of the brightest rays, that is, the yellow ; but this is 

 precisely the mean, double refraction, and besides, that of the 



