10 REPORT — 1866. 



arrives, on the contrary, at the expression 



tan a tan /3 



cos (i->r) - cos (i+,^-^°^ y- 

 It is obvious that oiu' formula agrees with tlie formula of Fresnel for the refracted 

 ray, and with the formula of M'^CuUagh according to the incident and reflected 

 rays. It is easy to conclude, from this theory, that imder the normal incidence 

 the luminous vibration rotates a right angle when the ray penetrates into the second 

 medium. It would be interesting to look for a direct veritication of that conclu- 

 sion ; l)ut it seems difficult to realize an experiment in which the surfaces limiting 

 the medium do not produce an even number of those rotations, so that the vibra- 

 tion does not come again to its first direction. The author could have stated 

 this property of polarized light under a more modest form, that is to say, as a 

 simple corollary of known theorems ; but he fancied that it was more useful, in the 

 actual state of optics, to state it as a new theory, in order to show, first, that the 

 geometrical simplicity of the principles does not constitute the most plausible 

 theories : thus it is prudent to conclude that the greater geometrical simplicit)' of the 

 M'^Cullagh tlieory is no sufficient ground for rejecting the theory of Fresnel, though 

 more complicated. Besides, the proposal of a new principle, very little obvious, u 

 priori, is a good occasion to remember the feeble degree of evidence for the prin- 

 ciples used in the other theoiies. After a further examination, it will appear that 

 it is neither more nor less difficult to admit that the refracted vibration is perpen- 

 dicular to both the others, than to admit, for the density of the luminous ether, 

 the theories of Fresnel or j\PCullagh. 



■On Dispersion-equivalents. By Dr. J. H. Gladstone and Rev. T. P. Dale. 



The refractive index of a substance minus unity, divided by its density, is termed 

 its " specific refractive energj'," and the product of this and its atomic weight, or 



P . , 18 termed its " refraction-equivalent." But the refractive index fi differs 



according to the part of the spectrum observed^ As the fixed lines A and H are 

 the extremes in the two directions which can be measured under ordinary circum- 

 stances, jx\i—y.\ has been taken as the measure of dispersion; and in a previous 



paper the authors had called ■ = — the " specific dispersion." Hence the dif- 

 ference between P -^-^^ — and P — y, — , or more simply P " , '^■, may be termed 



the " Dispersion-equivalent ; " and as P -^ is little affected by the manner in 



which the substance is combined with other bodies, it becomes a matter of interest 



to inquire whether the same holds true with respect to P ^h ^a . 



d 



It has been abundantly shown that the refraction-equivalent of the combination 

 Clio is 7-0 : its dispersion-equivalent, as determined from eight different substances 

 or series belonging to the great vinic group of organic compounds, varies only 

 within the limits of 0-32 and 0-38, the mean being 0-35. But when we turn to the 

 benzole group its dispersion-equivalent is found to be 0-62, and in the pyridine 

 group 0'58. 



The determinations of the dispersion-equivalents of chlorine, bromine^ and iodine 

 also differ, when they are made from different groups. 



Phosphorus is an extremely dispersive body, and when in a liquid condition 



gives for P ■ " "^ the high number 2-9, though the value of P for this element 



(I 



is low. 



Spectroscoj>e de pocl>e. Pur Dr. J. Janssen. 

 Get instrument est a vision directe, et forme une tres-petite lunette. Le redi'esse- 



