1867.] ( 277 ) 



11. PHYSICS. 



Light. — Some experimental researches on the indices of refraction 

 of saline solutions have been communicated by M. Le Verrier to the 

 Academy of Sciences. His aim has been to ascertain whether the 

 law, as laid down by Biot and Arago relative to gaseous mixtures, 

 was equally true for liquids. Having experimented upon 153 

 solutions, he arrives at the following conclusions, giving at different 

 temperatures for each body, its chemical composition, specific 

 gravity, and index of refraction. 



The number representing the index of refraction varies accord- 

 ing to the temperature. From 50° F. to 203° F. this variation 

 often attains the hundredth part, and is greater in proportion as 

 the liquid is the more concentrated. Its refringent power dimi- 

 nishes when the temperature is increased, and this diminution from 

 50° to 203° is about the thousandth part. The dispersion also 

 diminishes with the temperature. In the same interval the 

 distance between the two lines A and B of the spectrum may pre- 

 sent a difference of a thousandth part. The law of Biot and 

 Arago, that the index of a mixture of two gases is the mean between 

 that of the two elements which constitute it, is therefore not strictly 

 true in its application to liquid mixtures, but it may be regarded as 

 a very close approximation, differing little from the truth, for a 

 great number of saline solutions. 



When Nicol's prisms are used as polarizers or analyzers in 

 delicate optical measurements, an anomaly is frequently remarked : 

 the azimuths of extinction do not occur at a distance of 180°. 

 The error can amount to several tens of minutes. This error 

 would be fatal to the use of the Nicol's prism if the cause could 

 not be discovered, diminished, and remedied. M. Cornu first 

 pointed out this cause, and he has given the following explana- 

 tion : — The axis of rotation of the prism, or rather that of the 

 instrument which carries it, does not coincide with the plane of the 

 principal section ; hence the ray which traverses takes different 

 directions in the prism according to the azimuth, and the polariza- 

 tion to which it is subject is not parallel to the plane of the optical 

 symmetry of the crystal. When the lines of entry and emergence 

 of the prism are quite parallel, it can be regulated by trial ; in 

 general the error will be only alternated and not annulled ; but it 

 may be eliminated in proceeding by crossed observations. In fact, 

 it is easy to demonstrate by a very simple calculation and by direct 

 observation, that the error e of the normal azimuth is given by the 

 formula 



e = A (z + a); 

 A and e being the contents ; z the observed azimuth, it is easy to 



