on Magnetic Rotatory Polarization in Gases. 413 



third image, by alternating the measurements relative to the 

 red rays and to the green rays, produced: — 



Third image. 



Bed. 



Rotation 8*55 



Correction 3*19 



Difference... 5*34 



Green. 



9-61 



4-85 



4-76 



These numbers will be discussed later. From the begin- 

 ning of the experiments the series relative to the oxygen pre- 

 sented anomalies upon which I shall dwell presently. I also 

 made a further series of observations with this gas, some at 

 the beginning of the experiments in 1879, others at the end 

 of January 1880, and, finally, the determinations which I 

 have just recorded. 



The latter correspond to a purer gaseous mixture than the 

 previous series. I judged it useless to reproduce the numbers 

 relative to the former series ; they, however, completely con- 

 firmed those just given. 



Nitrogen. — The nitrogen was prepared by passing air very 

 slowly through two tubes, 0*50 metre long, filled with copper 

 turnings, heated to a red heat in porcelain tubes. The gas was 

 received in a large gasometer, and afterwards made to pass, 

 after drying, into the tube previously full of air. 



The gaseous mixture was analyzed by means of cold phos- 

 phorus; and out of 150 vol. of gas a residuum of 136*5 was 

 produced. The following are the results obtained: — 



Nitrogen, 0*91 + 0*090 Oxygen. 



Series. 



Average 

 tempera- 

 ture, t. 



Average 



pressure, 



H. 



Correction, 



(1 +atyiw 



Numbers 

 observed 

 and cor- 

 rections. 



Corrected 

 magnetic 

 rotations. 



Mean 

 devia- 

 tions. 



H 



White light. — 4th image (9 passages). 



Jan. 16, 1880. 



200 



millim. 

 762-6 



1-069 



11-00 

 Corr. 4-58 



Diff. 6-42 



I 6-86 



2-6 



The numbers obtained for air and oxygen ought to enable 

 us to find the number relative to the nitrogen. By correcting 

 the number above obtained by the 0*09 vol. of oxygen which 

 it contains, we obtain for the pure nitrogen the number 6*95. 



Let us now calculate the rotation of the nitrogen by means 

 of those of the air and oxygen. Let us take, for example, the 



Phil. Mag. S. 5. Vol. 12. No. 77. Dec. 1881. 21 



