116 Proceedings of the British Association. 



in order to restore the same system of rings, as e. g. that for 

 dark branches. These changes in general are analogous to, but 

 not the same as, those in their reflexion from transparent bodies 

 examined by Fresnel. But at the incidence for the maximum, as 

 well as at that nearest the perpendicular, they are the same. At 

 small incidences, in his former communication, the author mentioned 

 that he had been led to suppose an anomaly in this respect, the arc 

 appearing to deviate from 45°. But more recent and careful re- 

 petitions have shown that the results are really conformable to the 

 law in this case. He has carried on a considerable series of ob- 

 servations of the change of plane, and of the ellipticity, at different 

 incidences, and for various metals, &c. In regard to the maxi- 

 mum ellipticity, he has traced a relation to the constant arcs of 

 restoration after two metallic reflexions, determined by Sir D. Brewster 

 [Phil. Trans. 1830]. The change of plane cannot yet be explained 

 by theory ; though the empirical formulae of Prof. Maccullagh appear 

 to give a good representation of it; in the instance of steel, by 

 introducing the data resulting from that investigation into the author's 

 formula. 



This paper gave rise to a discussion, in which Dr. Lloyd, Prof. 

 Forbes, the Astronomer Royal, Mr. Challis, and Sir J. Herschel took 

 part. It having been stated in the progress of this, that Sir J. 

 Herschel had thrown out an opinion, that if the same amount of 

 analytic skill had been expended upon the corpuscular theory of 

 light as had been on the undulatory, perhaps more could be done 

 with it than was at present believed. — Sir J. Herschel said he had 

 not intended to bring any notions he might have casually thrown 

 out on this point before the Section, because in truth they were in 

 an indigested state ; but as the subject had been mentioned, what he 

 had said was, that physical differences of conditions could be con- 

 ceived, which might perhaps furnish the analyst with equations de- 

 pending on as few assumptions as those of the undulatory theory, by 

 which perhaps an explanation of the phenomena on the corpuscular 

 hypothesis might be worked out. Thus, for example, the individual 

 particles of light might be possessed of poles, and be thrown off 

 from the luminous body with revolving as well as progressive motions. 

 While these motions were uninterfered with, all the particles would 





