426 M. Verdet on the Optical Properties developed in 



of manganese and potassium, K^ Mn^ Cy^, which gave me a 

 remarkable result. This salt, when dissolved in water, so dimi- 

 nishes its rotary powei', that the salt itself must be regarded as 

 having a negative rotary power. Is the same the case with the 

 other salts of the sesquioxide of manganese ? T am not prepared 

 to answer this question decisively, but the excess of negative 

 power of persalts of iron over protosalts inclines me to answer 

 in the afBrmative. However this may be, and remaining strictly 

 within experimental limits, we see that manganese occupies in 

 one respect an intermediate place between iron on one side, and 

 nickel and cobalt on the other. That which is the rule with iron 

 salts appears to be the exception with those of manganese, and 

 vice versa. 



The interesting property of the double cyanide of manganese 

 and of potassium induced me to examine the double cyanide of 

 cobalt and potassium in the hope of meeting an analogous be- 

 haviour; but it was found to have a positive magneto-rotary 

 power of rather feeble amount. This double cyanide is even 

 diamagnetic. 



Chi-omium. — The salts of the protoxide of chromium are so 

 difficult to prepare, and especially to keep pure, that I did not 

 subject them to experiment. The salts of the sesquioxide have 

 so great a colouring power, that it is impossible to form an 

 aqueous or alcoholic solution containing a few per cent, without 

 destroying all transparence. But chromic acid and the chro- 

 mates are adapted for exact experiments. Neutral chromate of 

 potash, which is very soluble in water, gives clear yellow solu- 

 tions whose coloration admits of all requisite experimental exact- 

 ness, and only exercises a very insignificant influence on the 

 position of the passage-tint. Bichromate of potash, although 

 less soluble in water, gives solutions of deeper colour, but still 

 very transparent, and requiring no correction in determining the 

 position of the passage-tint. Chromic acid, on the contrary, 

 gives solutions of a very deep red, whose colour exerts a great 

 influence on the position of the tint, but which, in thicknesses of 

 from 1 to 2 centimetres, are sufficiently transparent. 



Solutions of these three substances have given me the follow- 

 ing results : — 



Table IV. 



