Mr. Potter on Conical Refraction. 351 



These results are certainly not in accordance with the theo- 



lions'of th""'""r' ^;^' ^"^'"^ H^-''^-' -'-- ^il 

 d cated bv Fr ''"S' ■ '•^^'•'''^.^•»"^/*^«'- the optic axes, as in- 

 dicated by Fresnel s expression for the wave-surface, have 

 be^,^ referred to in the quotations from Professoi Lloyd's 



The law of the polarization in the virtual ring and one of 

 fesso? no 'f r' T-}"' '"^.-^tigated it, were observed by Pro- 

 fesso Lloyd; but if my view of the mode of formation of the 

 ring by the intersection of two series of conicanurfeces be 

 correct, then there must be a peculiarity in the polarization 

 Imherto unsuspected. The ring, as stated by ProfSor Llovd 

 ispoarized in every part in a" different plLe -o that /we 

 attach a prism of calcareous spar, giving li.It no Wed^n 

 one plane, to our eye-lens, we see tlfat one smairpordon of 

 the ring vanishes in every position of the prism and S 

 we turn the prism and e^e^ens round the line of itn th" 

 dark part in the ring moves through the double of the ^nlla 

 which the prism does, making a whole revolution whil t fv, 

 pnsm has made the half of one. Now if the n^ I'lt 

 tersection of two series of cones, whilst one of"?he J'^^: 

 IS diverging and the other converging, the part which W? 

 pears, in any one position of the p.im sK L in the Z 

 sector measured from the optic axis as centre, in both serieT 

 but after one has ceased to converee the hri„l.r. . L ' 

 formed should exhibit no polarizado^n whatevef b ■^''' '^'" 

 posed of light polarized in^ all p^ars ;tr hen rclfveit^s- 

 S ' . P^'' ""u "'' disappears should be in the directly fn 

 posite sector to that which disappears in the .pwI f "^ ^' 

 which has continued to diverge. ^' ""^ ^^"^« 



This is exactly what we se^e; but where the two series nf 

 cones are diverging together, they soon overlap, and a tLv 

 are polarized in planes at right angles to each nfh.v ? ^ 

 point where they overlap, th^e JighftlLie exl bU n" ^7/. 

 polarization ; so that, for the internal series, it 1" onlv h"f 

 U as ^egun to overlap the other, that the pIiL J^ ,^ti! 



Sef^trlis-:,--^^^^^ 

 ^;f:^=:isxtr^^ 



of the virtual images within the crvs al now ! ) ' ''°" 

 have their distancts from the LZ? H.^ht y diff i:^^!^^ 

 pending, as we have seen, on their respectiC^ftacti e inWf 

 at that point, therefore the two portions o the '11^;: n,^' 

 different positions of greatest distinctness /amr^^lleiX' 

 seen, when experimenting in a dark room, ^ith a b gl pe'cj 



