546 FOOTE AND HARRISON: THERMOELECTRIC EFFECTS 



heating one of the wires near the junction, by drawing rapidly a 

 hot wire over a cold wire, by joining two wires of different di- 

 ameters and heating either wire near the junction, by filing a 

 groove or a v-shaped depression in a metal rod and heating this 

 portion of the rod, by passing a flame over a wire, and by numer- 

 ous other methods. The avoiding of these parasitic emfs has 

 made homogeneity testing a matter of some difficulty. 



All of these effects are well known and were discovered from half 

 a century to a century ago. Many of them are treated in Wiede- 

 mann's Lehre von der Elektriciidt, vol. 2. They are still of interest, 

 however, because conclusive evidence has not been given for the 

 causes of the various effects observed. Many reasons have been 

 suggested and have been supported by theory and experiment, 

 and no doubt among these many possible explanations one or 

 more are correct, but the absolute proof of their correctness 

 remains to be demonstrated. Thus, for example. Benedicks pre- 

 sents an experimental proof in which he makes use of a mag- 

 netic field. The conditions were such that very likely he 

 observed the well-known Nernst and von Ettingshausen effect 

 instead of the pure thermoelectric effect supposed. 



The priority for the discovery of these thermoelectric effects 

 probably belongs to Benjamin Franklin and Cavendish,- in 

 1769, or 147 years prior to Benedick's work. These investiga- 

 tors found that when a hot and a cold bar of the same metal 

 were placed in contact, the cold bar became positively charged, 

 as shown by measurements with an electroscope. 



The first evidence of the existence of a current in such a cir- 

 cuit was obtained by Rittei'^ in 1798. In the absence of am- 

 meters a pair of frogs was employed. The twitching of the legs 

 showed that positive current flowed in the circuit, frog leg — cold 

 Zn — (hot Zn-cold Zn) — frog leg. 



Since 1800 numerous observers have investigated the subject. 

 Indeed about 1850 these effects were thought by some, possibly 

 rightly, to be more fundamental than the Seebeck effect, and it 

 was believed that their study would lead to an interpretation 



2 Franklin and Cavendish. Experimental Observations on Electricity, p. 

 403. 1769. 



3RITTER. Gilb. Ann. 9:292. 1801. 



