186 Mr. R. Aclie on some Thermo-electrical Experiments. 



Again, with a bar of soft steel, and a similar bar made hard and 

 light by the well-known process of heating and immersing in cold 

 water, the thermo-electric action in this arrangement is from the 

 hard and light part to the soft and more dense portion of the 

 metal ; in this respect differing from the first couple, where the 

 current passed to the denser side of the two pieces of metal. A 

 couple formed of hard and soft steel elements affords another 

 good proof of the influence of the molecular state of a bar over 

 the thermo-electrical currents generated by the unequal heating 

 of it, the hardening of the steel being effected by heating and 

 quenching in water. The energy of the action of such a mono- 

 thermo-electric couple is very decided while the hardened steel 

 is undergoing the process of softening ; but after this has been 

 effected by the heat applied, the arrangement loses all its energy, 

 thus clearly showing a connexion between the molecular state of 

 the bar and its power to generate thermo-electrical currents. By 

 casting bars of antimony in hot and cold moulds, thermo-elec- 

 tric couples can be made of this metal which exactly resemble 

 steel in their action ; but they are not so uniform in their indi- 

 cations, the hardening and softening of antimony being much 

 less easily controlled. 



The metals bismuth and antimony occupy in thermo-electric 

 batteries the same relative position that zinc and platina do in 

 bydro-electric arrangements, — the bismuth corresponding to the 

 zinc, the antimony to the platina. When bismuth and antimony 

 are long employed in generating thermo-electricity, the antimony, 

 like the platina of the hydro-electric battery, is found unaltered, 

 while a minute change is always observed on the bismuth ; this, 

 then, points to bismuth as the active agent in developing the 

 thermo-electrical current, in the same manner as zinc is in the 

 hydro-electric couple. Bismuth is a metal possessed of some 

 remarkable properties, and it is to these that we should naturally 

 turn to examine the origin of the thermo-electric current. A 

 bar of bismuth and antimony soldered with bismuth only for a 

 solder, when put by the aid of a gas flame into as energetic action 

 as the fusing temperature of bismuth will admit of, develope 

 electricity, which if used to precipitate copper or silver after the 

 well-known manner of the electrotype, will after the lapse of a 

 few months give a precipitation equivalent to the weight of the 

 antimony and bismuth employed ; for this action a reaction must 

 be found somewhere ; all that the couple shows, is a minute 

 allotropic change of the bismuth where it has been in contact 

 with the antimony. When I first made this experiment, I was 

 inclined to believe that the small action in the joint was the 

 equivalent of the chemical effect produced at the electrodes ; but 

 subsequent reflection inclines me to doubt the accuracy of this 



