86 Dr. Tyndall on the Progress of the Physical Sciences : 



The two ends of the wire being connected with the galvanometer, 

 if either the hard side or the soft side be heated we have no 

 action ; but if one of the junctions of hard and soft be taken 

 between the finger and thumb, the heat of the hand is sufficient 

 to cause a deflection of 90 degrees. The writes has to thank 

 Prof. Magnus for an instrument of this kind. The wire presents 

 the same uniform appearance throughout j and to an observer 

 ignorant of the process to which the wire has been subjected, the 

 deportment is exceedingly striking and enigmatical. 



If two wires of the same material be taken, and if one be 

 heated and the other permitted to remain cool, on causing 

 the hot and cold wires to touch each other a current is observed. 

 This is modified if the one wire be hard and the other soft : 

 sometimes the difference of temperature and difference of hard- 

 ness work together and increase the current by their united 

 action ; sometimes they oppose each other, and a decrease of the 

 current is the consequence. This matter has been investigated 

 very fully. It will perhaps be well to describe beforehand the 

 manner in which the experiments were made. 



In a tin cylindrical vessel, AB, fig. 6, two tubes, ah and cd, 

 crossing each other at right angles, were introduced; each tube 

 had a diameter of half an inch ; from/, where the tubes crossed, 

 another vertical tube abutted upwards and passed through the 

 cover of the vessel ; the three tubes communicated with each 

 other inside ; through one of the horizontal tubes the wire to be 

 heated was introduced and fastened by corks at a and b ; to pre- 

 vent contact with the metallic vessel, all three tubes were lined 

 by smaller ones of glass ; at /' the wire was exposed, and rested 

 upon a flat piece of wood introduced beneath it ; in the vertical 

 tube was a wooden rod which nearly filled it, but could be moved 

 through the tube with freedom ; the rod carried at its end a 

 pound weight of lead, P ; the cylindrical vessel was filled with 

 water and kept constantly boiling, and as soon as it was certain 

 that the wire within had assumed the temperature of boiling 

 water, the wooden rod was raised, and the cold wire was intro- 

 duced crossing the warm one ; this being effected, the rod was 

 permitted to descend, and the wires were pressed together by the 

 weight P. The following table shows the results of this inquiry. 

 First both wires were hard, next both were soft ; and finally, the 

 one was hard and the other soft. 



