428 M. Verclet on the Optical Properties developed in 



metal, titanic acid is magnetic, and bichloride of titanium dia- 

 maffuetic*. 



Nevevtbeless I examined the magneto-rotary power of bi- 

 chloride of titanium, which is, as is well known, at ordinary tem- 

 peratures a transparent colourless liquid. It was found to be 

 negative, and gi'eater in absolute power than the magneto-rotary 

 power of water. 



Two points are noticeable in this experiment. In the first 

 place, it shows that the phsenomena under consideration depend 

 very little upon any chemical analogies which may subsist among 

 the several bodies. Chemists generally consider titanium as the 

 analogue of tin, and especially regard the chlorides of these two 

 metals as bodies of entirely comparable nature. On the other 

 hand, nothing is more dissimilar than the behaviour of the two 

 bodies when they are placed between the poles of an electro- 

 magnet, and examined with respect to their action on polarized 

 light. In the second place, by virtue of the liquidity and trans- 

 parence of the bichloride of titanium, its negative rotary power 

 may be shown directly, without the intervention of a solvent; 

 and thus an objection, which had been made to my former expe- 

 riments, is removed. Some persons considered the magnetic 

 action of certain iron solutions to be due to the fact that the 

 molecules of the magnetic compound in solution, being magnet- 

 ized by the electro-magnet, exercised a magnetic action contrary 

 to that of the electro-magnet itself, upon the neighbouring 

 atoms of the solvent. The experiments with solvents so varied 

 as those which I employed, and exercising, by themselves, so in- 

 considerable an action, appeai'ed greatly to invalidate this view ; 

 while my experiments upon chloride of titanium completely I'e- 

 fute it, by showing the existence of a diamagnetic liquid whose 

 rotary power is negative^. 



Cerium. — The magnetism of cerium was discovered by Faraday, 

 and is not more difficult to confirm than that of chromium or of 

 manganese. As I had no metallic cerium at my disposal, I ex- 

 amined two perfectly pure salts of this metal, a sulphate and a 

 chloride, which were prepared by MM. Damour and Deville in 

 the course of an investigation on cerium and the metals which 

 accompany it in its minerals J. Both salts showed a strong 



* In the note presented to the Academy of Sciences, July 8, 1857, I 

 declared myself unable to determine whether the bichloride of titanium 

 were magnetic or diamagnetic. By employing, however, the ingenious 

 process devised by M. Quct, for the study of the effect of magnetism on 

 liquids, I have been enabled to solve the question. 



t The objection in question was only made to me verbally, and has never 

 been mentioned in any publication relating to the subject of my researches. 



X MM. Damour and Deville had not determined the exact composition 

 of these two salts at the time when they furnished me with them. They 



