of the Plane of Polarization of Light. 311 



other metal began to deposit itself from the solution upon the 

 silver, the film came offthe glass in shreds. With a very good 

 film of silver and exactly adjusted strength of current it might 

 be possible to coat the silver with another metal electrolyti- 

 cally; at least J have at times succeeded in coating small 

 pieces of silvered glass with copper ; but I have not pursued 

 the attempt further, since a more convenient method presented 

 itself. M. Konig, in Paris, for the rotating mirrors employed 

 in his acoustical apparatus uses plate-glass, into one surface 

 of which a very thin transparent reflecting film of platinum 

 has been burnt. The film of platinum is so thin that it is 

 very transparent. Whether it is absolutely continuous is 

 not of any consequence ; at least it conducts so well that it 

 may be made an electrode, and covered with a reflecting layer 

 of another metal of any desired thinness. 



I have deposited iron according to Yarrentrapp's method * ; 

 cobalt and nickel from solution of the ammonio-sulphates of 

 cobalt or nickel. The concentration of the solution and 

 strength of current must be adjusted according to the thick- 

 ness which the film is desired to have. I have further coated 

 some of the iron and nickel films with reflecting surfaces of 

 gold, silver, or copper. These metallic films may easily be 

 obtained so thin that they, together with the platinum, are 

 quite transparent. 



The incident white light becomes brown when it traverses 

 the iron film, grey with nickel, and bluish grey with cobalt. 

 The surfaces of the deposited films usually reflect very well; 

 but the physical peculiarities of the different surfaces seem to 

 be somewhat variable. Possibly the metals are in certain 

 cases deposited in a crystalline condition. If we cut a long 

 strip from an iron film, it may easily be converted into a per- 

 manent magnet by rubbing with another magnet. It is 

 known that the coercive force of galvanically deposited iron is 

 very large. 



We are thus able to prepare permanent transparent iron 

 magnets, as also transparent magnets of cobalt or nickel. 



When suspended by a silk fibre, these transparent magnets 

 adjust themselves perfectly well in the magnetic meridian 

 under the influence of the earth's magnetism. 



I hope to study the behaviour of these very thin transparent 

 magnets later on. 



II. Rotation of the Plane of Polarization of Light by its 

 Passage through Iron, Cobalt , and Nickel. 



A Ruhmkorff's electromagnet of the usual construction was 



* Muspratt's Chemie, vol. ii. p. 1200. 



