218 Proceedings of the British Association. 



which it travels, an electric current can as well cool as heat the 

 point of contact of heterogeneous metals. He examined the effect 

 of friction on bismuth and antimony in the four following cases : — 



When heated. 



When at the temperature, 

 of surrounding space. 



When cooled. 



Antimony. Bismuth. 



Bismuth. Antimony. 



Bismuth. Antimony. 



Antimony. Bismuth. 



and he found, that the same act of friction produces in these different 

 cases, and with regard to the different metals, at one time a gain 

 of electricity accompanied by a gain of heat, at another time a 

 loss of electricity with a gain of heat ; again, either a loss or gain of 

 electricity joined to a loss of heat ; and asks, may it be hence inferred 

 that heat when nascent has a property specifically different from 

 that of heat residing in a metal ? Are we perhaps on the eve of 

 finding at length something analogous to the brilliant discovery 

 of Peltier, that galvanic electricity produces heat when proceeding 

 from antimony to bismuth ; and cold when travelling in the opposite 

 direction 1 by which M. Lenz has produced congelation. M. Erman's 

 chief object in communicating this paper to the British Association 

 was, to excite British philosophers to engage in this almost new 

 branch of investigation of tribo-thermic electricity. At the same 

 time, he excuses the imperfections of the investigations submitted, 

 by the extreme difficulties of such delicate experiments. Among 

 these, he particularized the almost unavoidable influence of hetero- 

 geneous rheophores, which must be joined to the metals under 

 examination. This difficulty caused him wholly to reject the first 

 set of his experiments, until at last he found that the interposition 

 of a plate of pure plumbago between each metal and its conduct- 

 ing wire proves an excellent means for limiting the anomalous pro- 

 duction of electricity, without lessening the conducting power of the 

 system. In conclusion, M. Erman suggests an important practical 

 application of this tribo-thermic electricity. Instead of the voltaic 

 apparatus in an electric telegraph, which is variable in its effects and 



