538 Mr. 0. J. Lodge on a Mechanical Illustration 



an extent proportional to this difference of potential, the elec- 

 tromotive force being balanced by the slight state of strain 

 which is set up in the air by this displacement. The metals 

 which show the greatest effect are those which stand furthest 

 apart in the voltaic series. 



But the contact electricity spoken of above as generated at 

 the junction of two metals has nothing to do with chemical 

 action. It is an incipient thermo-electric current taking- 

 place across the junction, which only shows itself as a slight 

 displacement of the cord, unless the metallic circuit is com- 

 pleted by a second junction generating an electromotive force 

 different from that of the first. The action between zinc and 

 copper would not be at all so strong as between bismuth and 

 antimony : in fact, the metals stand here in their thermo- 

 electric order. Of course, what is actually observed in any 

 ordinary case of two metals in contact is a mixture of the 

 two effects ; and the voltaic effect is usually far the greatest. 



§ 33. The very provisional value 2xr already obtained in 

 § 28 for the electromotive force at a junction is sufficient to show 

 several things. For instance, since it is a direct function of re- 

 sistance, substances of high specific resistances may be expected 

 to make the best thermo-electric materials — such, for example, 

 as metallic alloys, or sulphides, or, still better, tellurium or 

 selenium *. Hence also we might expect contact effects to be 

 extremely powerful between two substances ordinarily called 

 " insulators." Some violence might have to be used to dis- 

 place the cord through the very rough buttons of a dielectric ; 

 and a certain amount of heat would be generated by the 

 "friction" of the cord through the buttons; but when once 

 displaced it would not easily slide back again, and the effects 

 might accumulate ; so that by rubbing a dielectric with a 

 cord-supplying substance (a metal), one could electrify it very 

 highly ; and this is done in an ordinary glass electrical machine. 

 Any two substances which differ at all in conducting power 

 may be expected to give some contact-force. Thus, if a 

 crystalline metal conducts better along the fibre than across, 

 or if a strained metal has a higher resistance than an un- 

 strained, or if the conductivity of magnetized iron depends 

 on the direction of magnetization, then contact-force may be 

 looked for in each case at the junction of the substance in its 

 two different states. 



§ 34. It is now time to try what are the actual values which 



* Since the resistance of selenium has been found to fall off under ex- 

 posure to light, it is probable that its " thermo-electric power " will be 

 found to do so also ; in other words, illuminated selenium will have a 

 place higher up in the thermo-electric series than unilluminated selenium. 



