Physics. — ''(hl therwoelectric airnmls in mtrcnrif'. Hv Prof. 

 H. Hag.\ and Dr. F. Zernikk. 



(Communicated in the meeting of March 29, 1919). 



In an extensive paper "Ein tnr Tlierino-electrizitat nnd metallische 

 Warmeleitnng- fiindanientaler P^ffect", ^) C. Bknedicks endeavours to 

 show the existence of a thermoelecti-ic force in a iioniop;eneous con- 

 ductor, when at both sides of a heated part the temperature falls off 

 at a different rate, i.e. the temperature gradient is different. ') He 

 makes due allowance for the possibility of the thermo-electric force 

 which appears under these circumstances in solid conductors l)eing 

 due to changes of structure, which would make the contiguous parts 

 of the coiuluctor behave as different substances. Therefore he 

 considers the proof of the reality of the force mentioned to depend 

 upon the success of experiments with a liquid conductoi. To this 

 end he used a glass tube filled with mercury, which had beeji drawn 

 dowMi at one point, and heated the wider part next to it, thus 

 bringing about a high temperature with slow decrease in the wide 

 part, and steep gradient at the constriction. He did not obtain 

 reliable results from these experiments, which he supposed to be 

 caused by the conduction of heat through the glass. This would 

 heat the mercury in the narrow part and thus prevent the tempe- 

 rature from falling ofi' with sufficient steepncis. Against the 

 well-known experiments of Magnus, who coidd not detect any 

 current in bringing into contact mercury conductors of different 

 temj)erature, and who worked also with unequal temperature gradients, 

 l)EiNEDicKs advances the low sensitiveness of Magnus's pointer-galva- 

 nometer compared with the mirror galvanometers of the present time. 



We have devised an experimental method by which, while 

 using mercury, all the conditions of Benedicks for a sensitive test') 

 would be met, thus enabling us to get a conclusive decision as to 

 the reality of the new effect. This method consists in the use of 

 two mercury jets joined to a sensitive galvanometer, one of them 



1) Ann. der Physik. 55. p. 1—80, p. 103—150. 1918. 



2") Pogg. Ann. 23, p. 497, 1851. Magnus here tried to disprove the production 

 of thermo-electric currents under these circumstances. 

 s) I. c. p. lis. 



