A Netif Thermoelectric Phoiovieium. 120 



Several other tubes were filled with tin and heated, and with one 

 exception all gave from -1 to -2 volts, from the remaining one 

 only infinitesimal results could be obtained. The changes of 

 e.m.f. were much slower in tin than in lead, but it was very 

 sensitive to the flame being shifted along the tube, a few milli- 

 metres of shift sometimes causing a great variation in the galvan- 

 ometer reading. There were no abrupt changes as in the case of 

 lead on removing or lowering the flame. A tin wire (alloyed 

 with lead) when put through a tube and heated behaved similarly 

 to tin and lead, • 1 6 volt was obtained from it. 



I could not get a tube filled with zinc in the same 

 way as I had filled others with lead and tin, but I 

 managed to get one filled by exhausting it while the 

 end was dipped into a crucible of melted zinc. The highest 

 e.m.f. I observed with this was only -00035, but that was 

 partly, if not wholly, due to the hot junction of the zinc and 

 copper, as the tube of zinc was very short. I afterwards got some 

 zinc wire and passed it along a tube and heated it. There was 

 no result at first, and the wire fused and broke circuit as the 

 diameter of the wire was much less than the bore of the tube. 

 When the ends were pushed in and contact renewed -05 volt was 

 indicated, but it quickly fell to '004. On shifting the flame 

 there were various smaller effects, but after cooling and heating 

 several times -2 volt was at length reached, the behaviour not 

 being in any way characteristic. At a time when the e.m.f. was 

 •01 and falling slowly I turned ofi" the gas. It fell somewhat 

 faster, though still slowly, and after some minutes, when the tube 

 was cool enough to be held in the fingers, -006 volt was still 

 indicated. The temperature of the zinc could not have been over 

 200\ 



As already mentioned I had examined copper to see what 

 efiect could be obtained from it and had only reached -0001 volt. 

 After obtaining such high eflects in other metals I returned again 

 to copper, using the finest wire I could get, this was -16 mm. 

 diameter, and silvered, but the silver disappeared almost instantly 

 on heating. After a little irregular heating, '001 volt was 

 reached, the changes being very slow, and oscillations of the 

 needle being scarcely perceptible. After some time, however, 



