548 Intelligence and Miscellaneous Articles. 



identical with that representing the distribution of magnetism in a 

 magnetized bar : — 



reckoning os from the middle of the plate ; or 



counting from the margin. The differences between calculation 

 and observation were always less than the possible errors of obser- 

 vation, and only exceptionally rose to O'Ol of the measured length. 



I will add that, in a long rectangular plate, for the distribution 

 of the fringes parallel to the longer sides a curve was obtained 

 similar to that of brachypolar magnets ; any square plate gives a 

 similar curve to that of megapolar magnets. For example, there 

 are found : — for a rectangular plate, A =2-53, K = 2*99 ; for a square 

 plate, A=O0124, K= 100-5. The ordinate-unit is the half of a 

 wave-length ; the unit of length, the centimetre. 



After these researches the question naturally arose, Are these 

 phenomena to be attributed, as Neumann thought, to a regular 

 action exerted upon the glass mass by the outer layers at the time 

 of its solidification ? In regard to this I have made several series 

 of experiments, varying the conditions ; but in all cases I was con- 

 ducted to the same conclusion : the causes of these two phenomena 

 are completely different. 



It is quite otherwise when the effects of tempering are compared 

 with those resulting from regular heating of a plate of glass through 

 its contour. Examining with the polarizing microscope a square 

 plate of glass heated by the usual processes, one is struck with the 

 similarity of the phenomena. Unfortunately, the rapidity with 

 which the colorations change and disappear did not permit precise 

 measurements to be made ; I was therefore obliged to have recourse 

 to an artifice. 



I took two rectangular plates derived from the same plate 

 of glass, and both having absolutely identical dimensions. One 

 was feebly tempered ; the other could be introduced into a hot 

 metallic piece constructed specially for the purpose. It was 

 at once perceived that the signs of a tempered and of a heated 

 plate are the same. If, moreover, after arranging upon the pola- 

 rizing microscope the tempered plate at 45° from the plane of polari- 

 zation, the healed plate of the same shape be superposed'crossing 

 it, a moment comes when the common part is traversed by a 

 black cross, the arms of which pass through the four vertices 

 of the square of intersection. This phenomenon can only be 

 produced if the distribution of the retardations is at that instant 

 the same in both plates. Double refraction produced by tempering 

 is therefore identical with that which is produced by a regular 

 heating through the contour of the plate. — Corhptes Mendus de 

 V Academic des Sciences, May 7, 1877, tome lxxxiv. pp. 1024-1026. 



