424 M. Verdet on the Optical Properties developed in 



my second memoir*, gave rise to rotations of too feeble a kind : 

 with a thickness of 3 or 4 centimetres the solution was not suffi- 

 ciently transparent for accurate observations. 



Nitrate of the peroxide of iron, subjected to the action of mag- 

 netism, exerts an action on polarized light in the same direction 

 as the perchloride, but of less energy. An aqueous solution of 

 this salt has a less magnetic power than water : when the solu- 

 tion is quite concentrated, the rotary power is almost nothing, 

 but there is no change of sign in the rotation. 



The two prussiates of potash, or double cyanides of iron and 

 of potassium, seemed to me to be worthy of a special study. In 

 fact, we know from the experiments of Faraday f and Pliicker J, 

 that the yellow prussiate is diamagnetic, and the red prussiate 

 magnetic. I have found that the magneto-i'otaiy power of the 

 yellow prussiate is positive and moderate, while that of the red 

 prussiate is negative and very great. 15 parts of the red prus- 

 siate dissolved in 85 parts of water, give a liquid whose magneto- 

 rotary power is negative, and in absolute value nearly double 

 that of water. Inasmuch as it will be seen in the sequel that 

 there are diamagnetic compounds of magnetic metals whose 

 magneto-rotary power is negative, it is probable that the posi- 

 tive power of the yellow prussiate is not owing to the diamag- 

 nctism of this compound, but to the fact that the physical pro- 

 perties of the iron it contains are as completely masked as the 

 chemical ones. 



Nothing is easier than to prepare an aqueous solution of per- 

 chloride of iron or red prussiate of potash, suitable for showing, 

 even on the lecture-table, the negative action of iron salts on 

 polarized light. Nevertheless it would be advantageous for this 

 purpose, and interesting for other reasons, to possess solid trans- 

 parent bodies having the same properties as solutions. I must 

 confess that my search for such bodies has hitherto been fruit- 

 less. Some very transparent crystals of iron-alum, 1 centim. in 

 thickness, given me by M. Deville, showed so high a degree of 

 lamellar polarization, that I could make no use of them. Several 

 fragments of iron-glass which I tried, were either not sufficiently 

 transparent, or not sufficiently free from flaws. I hoped to be 

 more fortunate with the glasses of phosphate of lime, which M. 

 Margueritte has recently shown to he capable of taking up the 

 most varied metallic oxides. I\I. Margueritte was kind enough 

 to have several pieces made for me, containing from 5 to 20 per 

 cent, of peroxide of iron. But none of these pieces, even after 

 a very prolonged re-annealing, were found sufficiently free from 

 flaws for experimenting on. It is probable that among the 



* See Annales de Chirnie et cle Physique, 3 scr. vol. xliii. p. 3/. 



t Pliil. Trans, for 18-lG. J PogguudoriT's Annalen, vol, Ixxiv. 



