248 Dr. J. Kerr's Electro-optic 



to return to its original form ; and so we froze water in two 

 f-inch iron gas-pipes, one round and one which had been flat- 

 tened when red-hot. The round pipe burst the first time; 

 the flat one did return slightly when thawed, but not enough 

 to prevent its bursting during the second operation. 



No doubt most people will not consider this proposal of 

 Mr. Powell's a satisfactory cure for burst pipes : they would 

 like something which could be fixed in their houses and which 

 would be always safe without further attention. But till such 

 a discovery is made, I think elliptical pipes give the best solu- 

 tion of a problem which has troubled every householder. 



As the subject of this paper is of physical as well as general 

 interest, I hope that it may be considered not unworthy of the 

 attention of the Physical Society. 



XXXI. Electro-optic Experiments on various Liquids. By 

 John Kerr, LL.D.,Free Church Training College, Glasgow. 



[Concluded from p. 169.] 

 Su Iph ides, ( CJEIa, + 1) 2 S. 



13. nnHESE are purely negative, and rise in power as a 

 -L increases ; they are weaker than the corresponding 

 oxides, but much stronger than the hydroxides. 



Ethyl sulphide, not a clean liquid, required the jar and 10 

 to 20 turns of the plate. The effect was purely negative, 

 strengthened by compression parallel to lines of force, and 

 extinguished by tension. 



Butyl sulphide, tried as a nonconductor, gave a very faint 

 effect, which was barely characterizable as negative. Tried 

 then as a conductor, with increasing charges of the jar, it gave 

 a set of brilliant effects, always purely negative. 



Amyl sulphide, a feeble photogyre, hardly separating the red 

 and blue at extinction, was rather stronger than the last. It 

 acted well as a nonconductor, giving a clear effect which was 

 purely negative. As a conductor, with increasing chai 

 gave increasingly brilliant effects, always negative by both 

 tests. 



Ethyl dmdphide was also examined, and was found to be 

 purely negative, like the protosulphide, but much stronger. 

 As a nonconductor it gave a pure and moderately strong effect; 

 and as a conductor, with increasing charges of the jar, it gave 

 a fine series of restorations, all purely negative, extinguished 

 perfectly by tension parallel to lines of force. 



