SCIENCE AND MAN. 345 



duce decomposition, magnetise iron, and deflect a mag- 

 netic needle at any distance from its origin. You will 

 be disposed, and rightly disposed, to refer those distant 

 manifestations of power to the heat communicated to 

 the face of the pile, but the case is worthy of closer 

 examination. In 1826 Thomas Seebeck discovered 

 thermo-electricity, and six years subsequently Peltier 

 made an observation which comes with singular felicity 

 to our aid in determining the material used up in the 

 formation of the thermo-electric current. He found 

 that when a weak extraneous current was sent from anti- 

 mony to bismuth the junction of the two metals was 

 always heated, but that when the direction was from 

 bismuth to antimony the junction was chilled. Now 

 the current in the thermo-pile itself is always from bis- 

 muth to antimony, across the heated junction a direc- 

 tion in which it cannot possibly establish itself with- 

 out consuming the heat imparted to the junction. This 

 heat is the nutriment of the current. Thus the heat 

 generated by the thermo-current in a distant wire is 

 simply that originally imparted to the pile, which has 

 been first transmuted into electricity, and then retrans- 

 muted into its first form at a distance from its origin. 

 As water in a state of vapour passes from a boiler to a 

 distant condenser, and there assumes its primitive form 

 without gain or loss, so the heat communicated to the 

 thermo-pile distils into the subtler electric current, 

 which is, at it were, recondensed into heat in the dis- 

 tant platinum wire. 



In my youth I thought an electro-magnetic engine 

 which was shown to me a veritable perpetual motion 

 a machine, that is to say, which performed work with- 

 out the expenditure of power. Let us consider the 

 action of such a machine. Suppose it to be employed 

 to pump water from a lower to a higher level. On ex- 



