772 ANNUAL EEPORT SMITHSONIAN INSTITUTION, 1908. 



he needed an idea of the size of the eft'ect which he might expect. 



R 



The preceding hiw, ., , ., -. y = 0.2, approximately, used with gases 



of which the refractive indexes are known, shows that the amount of 

 rotation should be very different from one end of the spectrum to 

 the other and of the order of 0.0001 of- that due to carbon bisulphide. 

 So with an apparatus having a magnetic field 30 meters in length 

 he expected a rotation of five to ten minutes of arc. Faraday had 

 proved that the effect may be increased by repeated reflections of the 

 light across the magnetic field. Becquerel employed this device in 

 an apparatus about 3 meters in length to measure the rotary power 

 of gases and their relatively great rotary dispersion. 



Oxvgen was found to be abnormal. Its rotary dispersion is ex- 

 tremely small and perhaps anomalous, the green giving less rotation 

 than the red. This behavior may be compared with the well-known 

 magnetic properties of oxygen. Carbon bisulphide furnishes strong 

 evidence in support of the law, for in the liquid state the constant 

 is 0.231 ; in the gaseous, 0.234. 



Becquerel naturally searched for applications of his theory in the 

 world about him. He looked for the action of the terrestrial mag- 

 netic field upon light, arid with carbon bisulphide observed deviations 

 of the plane of polarization of about half a degree. But a new ques- 

 tion arose, Does the terrestrial magnetic field deviate the plane of 

 atmospheric polarization, which, according to theory, must pass 

 through the center of the sun ? This study, requiring measurements 

 of high precision, showed that the plane of polarization undergoes a 

 daily oscillation about the theoretical plane, due for the most part to 

 diffuse light, but there is a small residual variation caused by the 

 earth's magnetism, 150 kilometers of air giving a rotation of about 

 20' of arc. 



This elaborate series of experiments would not have been complete 

 had the practical side been neglected. At the congress of electricians 

 in 1881 Becquerel proposed to measure electric currents in absolute 

 units by means of the rotation of the plane of polarization by carbon 

 bisulphide. He showed that this method is free from the perturba- 

 tion at the ends of the magnet, the rotation formula being 

 E=K.4 IT NI. By various experiments, made by the deposit of silver, 

 the constant, K, w^as found equal to 0'.04841 at 0°. This gives a 

 method of constructing a secondary scale for the ampere, precise 

 though difficult in practice. 



To return to the theories, which had furnished motive for all these 

 experiments, Becquerel combined his ideas in a theory of the mag- 

 netic rotary polarization and showed how analogous results could be 



