Intelligence and Miscellaneous Articles. 547 



is continually changing its values, from the rest which remains 

 constant. It is therefore likely that the classification of light into 

 three kinds, according as the violet, the red, or the green is contained 

 in the smallest proportion, is one which has a relation to the natural 

 powers of discrimination. — Silliman's American Journal, April 1877. 



ON ACCIDENTAL DOUBLE REFRACTION. BY J. MACE. 



The phenomena of accidental double refraction produced by com- 

 pression have, since their discovery by Brewster, given rise to 

 numerous researches, while those produced by tempering have been 

 but little studied. Since the discovery of the latter by Seebeck, we 

 find but little concerning them except the papers of M. de Luynes, 

 in which the author studies chiefly the mechanical properties 

 acquired through an energetic tempering, and those of M. Maseart 

 on the tempering of cast plates of glass {Journal de Physique, 1876), 

 in which the author applies himself especially to the practical side 

 of the question. 



The aim of the investigation which I commenced some months 

 since was, on the contrary, to discover what are the laws which 

 govern this phenomenon, and to search out the analogies or the 

 differences which might exist between that which results from tem- 

 pering and the other cases of accidental double refraction. I 

 therefore proposed to myself to study plates of simple geometrical 

 forms, especially rectangular and square, differing in dimensions, 

 thickness, and constitution. If it be remarked that "Wertheim, by 

 the study of the regular compression of glass, has been able to show 

 that the differences in the course are proportional to the pressures 

 exerted, it will be seen that this led me to investigate the distribu- 

 tion of the differences of the course in the various plates submitted 

 to experiment. 



The arrangement employed was very simple : — A car carrying 

 the plate permitted it to be displaced parallel to a micrometer 

 formed, as in M. Jamin's apparatus, of two parallel fine threads and, 

 besides, of a third, horizontal thread, rendered necessary by the 

 curved form often (and particularly in square plates) affected by the 

 fringes to be studied. As always, the polarizer was placed at 45° ; 

 the analyzer was alternately transverse and parallel, so as to exhibit 

 fringes corresponding to differences of course varying by half wave- 

 lengths." The compensator was suppressed, it having the inconve- 

 nience of deforming the fringes and often rendering them difficult 

 to observe ; but it was made use of for the rectangular plates, in 

 order to measure the central difference of course. For the illumi- 

 nation the burner of Laurent's saccharimeter was used. The posi- 

 tion of each fringe was measured to within 0*1 of a millimetre at 

 the least. 



In studying the distribution of the course-differences along a line 

 parallel to and equally distant from two of the sides of the plate, 

 the phenomenon could in all cases be represented by a formula 



