464 Mr. G. W. Wiilker on the 



nature of the charge after its penetration into the mass of the 

 dielectric. 



Conclusions. — All that can be said at present of these acces- 

 sory phenomena is that they do not permit us to affirm -svith 

 certainty that the third fundamental condition of a correct 

 experiment on convection is exactly fulfilled, particularly 

 with sectored disks. But the following conclusions are 

 legitimate : — . 



1. A charo-ed continuous metallic disk turnino- in its own 

 plane opposite fixed conden sing-plates carries its charge 

 with itself. 



2. The entrainenient of this charge produces a magnetic 

 field in the direction demanded by the assumption of a 

 magnetic eifect due to electric convection, and in accord with 

 the calculated value to 10 per cent. 



3. Charged sectors moving in their own plane, without the 

 presence of any condensing-plates, produce a magnetic effect 

 in the direction and of the proper size demanded by this 

 same assumption. 



It is not for us to say whether these etfects are really due 

 to electric convection in the sense in which Faraday and 

 Maxwell understood this expression, nor to decide whether 

 they are in accord with the fundamental hypotheses of the 

 accepted theories. 



Plivsical Laboratory of tlie Sorbomie, 

 " Paris, April 15, 1903. 



LIX. On the Theory of Refraction in Gases. By Geoege 

 AY. Walker, M.A.^ A.B.C.Sc, Fellow of Trinity College, 

 Cambridge'^. 



[Plate XIX.] 



THE problem involved in the explanation of the properties 

 of refraction and dielectric susceptibility in gases, has 

 occupied n^y attention for several years. The subject is of 

 extreme importance on account of the evidence which a 

 satisfactory theory could give of the constitution of the mole- 

 cules of a ^as. 



More than two years ago I suggested a theory of refraction 

 in gases t; but while the present theory is an extension of 

 the former, careful consideration of more experimental 

 evidence has led me to modify some of the views expressed 

 there. 



Several theories in terms of the electromagnetic theory of 

 light have been proposed ; but while these explain some of 



* Communicated by the Author, 

 t Proc. Roy. Soc. vol. Lxix. p. 394. 



