[ 370 ] 



LIII. On the Motion of Luminous JVaves in an Elastic Me- 

 dium, consisting of a system of detached particles^ separated. 

 by finite intervals. By S. Eaunshaw, M.A.^ of St. John's 

 College, Cambridge'--. 



T RECEIVE with distrust results on the subject of the 

 * transmission of waves of light through a medium of de- 

 tached particles, when those results are derived from investi- 

 gations based in any degree upon an assumed geometrical 

 symmetry in the arrangement of the molecides of the medium. 

 That there exists in nature a mechanical symmetry amongst 

 x\\e forces which regulate the motions of the particles of the 

 luminiferous medium, is proved by innumerable experiments: 

 for, in those media which are denominated uncrystallized, 

 the forces are symmetrical with respect to every direction ; 

 in those which are denominated uniaxal the forces are sym- 

 metrical with respect to a certain line, and to all lines at right 

 angles to it; and in biaxal media there exist three rectangular 

 directions, with respect to which there is mechanical sym- 

 metry. Now if we consider the circumstances of the last two 

 classes of media, we shall perceive that the directions of me- 

 chanical symmetry are very rarely directions of geometrical 

 symmetry; and that even when there are no lines of geome- 

 trical symmetry, there yet exist three lines of mechanical 

 symmetry. With these facts before me, well estabHshed by 

 almost innumerable experiments, I cannot read with satisfac- 

 tion the researches of those mathematical philosophers who 

 make an assumed (may I not even say an uufmmdedl) geo- 

 metrical symmetry the basis of their investigations. If we 

 inquire into the reason of this fondness among writers on the 

 subject of physical optics for this imaginary geometrical sym- 

 metry, we shall find that it is forced upon them by the com- 

 plex nature of the equations which present themselves in their 

 investigations; and it is used as a means of simplification, 

 without which there seems to be no hope of obtaining any 

 result at all. I shall show presently that simple results can 

 be obtained, when the investigations are conducted on proper 

 hypotheses, warranted by experiment, without the assistance 

 of this means of simplification. 



I am not prepared to deny that, to a certain extent, geo- 

 metrical symmetry may secure mechanical symmetry; but, as 

 remarked above, experiment shows that the latter exists in 

 every system whether the former exist or not; and there- 



* Communicated by the Author. 



