128 Messrs. Gee and Holden on 



body belongs to Class II. *; if, on the other hand, both W p , and 

 E w are partially discharged, B is placed in Class III. If 

 E p alone is discharged, B is a positive unipolar conductor; if 

 this is the case only with E n , B is a negative unipolar con- 

 ductor. 



At the time when Erman classified conductors in this 

 manner Ohm's law was not known, hence the behaviour of 

 the electroscopes under these conditions could not be thoroughly 

 appreciated. However, it was recognized that in the cases of 

 Classes IV. and V. there must be a high resistance at one of 

 the contact points n or p, preventing the conduction of the 

 charge of E n or E p to earth. 



It was found that well-dried soap was a good example of 

 Class V. Hence Erman studied the unipolar conducting- 

 power of soap in some detail, but he was unable to come to 

 any definite conclusion why at p there should be a very high 

 resistance. He rejected, on grounds which need not here be 

 described, the hypothesis of an insulating film composed of 

 the decomposed soap- acids j\ 



Interesting examples of Class IV., Erman found, are fur- 

 nished by flames. Their anomalous conductivity yet remains 

 but imperfectly explained J. 



2. Experiments of Ohm. (a) On Soap. 



Ohm § confirmed the experiments of Erman, and came to 

 the conclusion that the cause of the high resistance at the 

 anode was due to a layer of fatty acids. 



* The following extract from Erman's paper (loc. cit. p. 18) is interest- 

 ing, as evidence of his scientific caution : — " The distinction between 

 perfect insulators and perfect conductors was perceived very early with 

 the battery, and though it may indeed be believed that no important dis- 

 coveries remain to be made in this matter, it is very desirable not to con - 

 sider the matter as fully settled, for it is possible that also here important 

 modifications of conductivity lie hidden. For example, it has really not 

 yet been proved whether, with the so-called perfect conductors, there are 

 not degrees of conductivity. Thus, though it is usual to regard all metals 

 as equally good conductors, I should like this law not to be so uncon- 

 ditionally assumed, before direct measurements have proved its generality. 

 Indeed, who knows whether finally with each metal certain properties 

 will not be made apparent, which in respect to galvanic electricity are 

 analogous to those which so excellently characterize iron, nickel, and 

 cobalt with regard to magnetism ?" 



t Loc. cit. p. 50. 



% The conducting power of flames has been studied by Ritter, Hittorf, 

 Andrews, Hankel, Braun, Fuchs, Macfarlane, Rintoul, and others. 



§ Schweigger's Journal, lix. (1830) p. 385. 



