522 



On the Chloride -of- Silver Battery. [Dec. 13_, 



culty of keeping them absolutely parallel when the micrometer is 

 adjusted from time to time. 



The striking- distance between two paraboloidal points was found to be 

 with : — 



1080 cells. 8040 cells, 



in. in. 

 0-005 0-401 



The nature of the metal used for terminals has, in almost all cases, no 

 influence on the length of the spark ; but there is one striking exception, 

 namely, in the case of aluminium. When an aluminium point is used the 

 spark is longer than with points of all other metals tried in the ratio of 

 1-242 to 1. 



The length of the spark is different in various gases — for example, air, 

 oxygen, nitrogen, hydrogen, and carbonic acid ; and the ratio between the 

 lengths of spark in various gases varies with the forms of the terminals. 

 The length of the spark bears no simple relation either to the density of 

 the gas or its viscosity*. 



The paper contains an account of a few experiments on the length of 

 spark in air at different pressures, from 141-5 millims. to 760 millims. 

 Between a point and a disk the length of the spark increases nearly, but 

 not quite, in the ratio of the dilatation ; but between two spherical sur- 

 faces it increases far more rapidly, and it is possible that at a certain 

 degree of rarefaction the striking-distance may be coincident for spherical 

 surfaces and points. 



The appearance of the voltaic arc at ordinary pressures differs in 

 different gases. 



In air (possibly also in other gases) the arc, when examined with the 

 microscope, presents a stratified appearance, especially in the barrel- 

 shaped surrounding of the central brilliant spindle. The striae are very 

 close, and can be seen with difficulty even when the microscope, with a 

 rotating minor, is employed for the examination of the arc. 



In hydrogen, with the point positive, the central spindle of the arc is 

 surrounded with a beautiful blue halo resembling a glass shade illumi- 

 nated with fluorescent light. "With the point negative the arc moves 

 about rapidly and forms a sort of star on the positive disk. Before the 

 jump of the spark, when the point is negative, the luminous discharge 

 has the appearance of a pale olive halo, inform like a glass shade, extend- 

 ing from the point to the periphery of the disk. 



In nitrogen the arc is reddish violet ; in oxygen it presents the same 

 appearance as in air. 



When a strong resistance is interposed in the circuit, 4,000,000 ohms 

 for example, the discharge is completely changed in character; instead of 

 the ordinary spark and production of the voltaic arc, very brilliant snap- 



* Proc. Roy. Soc. vol. xxvi. p. 227. 



