1804.] On the Electrical Properties of Pure Substances. 39 



Part I is then continued. 6 deals with an investigation of the 

 specific inductive capacity of various kinds of sulphur by the method 

 of weighing, advantage being taken of the laborious investigation 

 of the method dealt with in Part II. Various other matters came to 

 light, and we furnish a drawing of suitable apparatus and describe 

 the necessary course of procedure to make the method accurate and 

 satisfactory ; in particular the proper way of preparing plates of 

 crystalline and friable substances. The results for the specific induc- 

 tive capacities are as follows, at a temperature of 14 C. 



" Aged " monoclinic sulphur K = 3'162 



Ditto with 1'43 per cent, insoluble unstable 



sulphur K = 3-510 



Ditto with 3 per cent, insoluble unstable 



sulphur K = 3-75 



An experiment on purely amorphous sulphur is not yet ready for 

 publication ; but the above results will go a long way to clear up the 

 great differences in the hitherto published values of this constant 

 for sulphur. They also serve as a check on our observations on thin 

 films, and show that our measurements of film thickness a grave 

 difficulty were moderately successful. 



A number of experiments bearing on a theory of conduction which 

 we venture to suggest are also included in this part of the paper. 



This is followed by an account of the theory to which our experi- 

 ments led us, and which is briefly as follows. Sulphur in either of 

 the extreme conditions does not conduct ; we can only examine the 

 purely soluble state, for the other is not sufficiently stable for a 

 satisfactory investigation ; however, we may say that changing the 

 content of amorphous unstable sulphur from 3 per cent, to 88 per 

 cent., produces little or no change in the conductivity. Taking this 

 and other facts into consideration, we believe that what we have 

 called mixtures of the two kinds of sulphur are really compounds, 

 and that the conduction is electrolytic. 



We have framed what we believe to be a novel theory of electro- 

 lysis, which explains all the facts which we have observed, and which 

 has the peculiarity of introducing the idea of an electrolytic con- 

 vection current, in connection with which the resulting changes of 

 specific inductive capacity allow of all the phenomena of conduction 

 observed taking place, though the charges may never really reach 

 the electrodes. It will be seen that the effects of fatigue-reversal 

 and the phenomena of discontinuous conduction are well accounted 

 for by this theory. The only objections we have to it are that it is 

 based on a molecular theory of matter, which we are persuaded requires 

 to be remodelled, if it is to afford any real explanation of things as they 

 are. A theory of residual effect based on the theory of conduction is 



