on Magnetic Rotatory Polarization in Gases. 313 



arrangement of the apparatus is shown in the accompanying 

 figure (Plate VIII. fig. 1). 



2. Optical Apparatus. 



Under the conditions of magnetic intensity just described, 

 the rotation of the plane of polarization of a luminous ray 

 passing through the tube filled with gas is very feeble. With 

 the various gases studied, the double rotation obtained by 

 reversing the direction of the electric current in the bobbins 

 did not exceed an angle of h' . On this account it became 

 necessary to amplify the phenomenon. Two methods might 

 be employed. The first, contrived by Faraday, consists of 

 passing several times through the substance to be studied 

 luminous polarized rays which are reflected from two mirrors 

 conveniently placed. It is well known that, under these con- 

 ditions, the magnetic rotation is proportional to the distance 

 traversed through the substance influenced, contrary to what 

 would be observed in a body endowed with a natural mole- 

 cular rotatory power. This method is the one exclusively 

 adopted by me in the experiments about to be described. 



The second method is a general method of amplification of 

 the displacements of the plane of polarization of a luminous 

 ray, and was proposed and applied by M. Fizeau. It consists 

 in passing the polarized luminous ray through a pile of glasses 

 conveniently diverged. I have tried to apply this method; 

 and I shall state hereafter why, in the present instance, I have 

 preferred to use the former. 



Description of the Optical System. 



The apparatus was specially constructed with a view to ap- 

 plying Faraday's method of amplification. 



Source of Light. — The source of light was incandescent 

 lime produced by an oxyhydrogen flame, and seen either 

 directly or through different-coloured screens. I employed as 

 screens coloured glasses — red, yellow, and copper-green — and 

 also a liquid screen containing ammoniacal nitrate of copper, 

 w r hich chiefly permitted light-blue rays to pass. We shall see 

 later how it was possible to estimate the length of the average 

 wave of luminous rays which in each case reached the eye of 

 the observer. 



The luminous point L (see PI. VIII. fig. 2) was placed at about 

 0'487 metre in front of the extremity of the tube — that is to 

 say, at 0*627 metre from the face of the last bobbin — at the 

 focus of a collimating-lens C, having a focus of 014 metre, 

 which transmitted through a polariscope P parallel rays. The 



