Svanberg on the Monothermic Currents. 89 



In the case both of bismuth and antimony the bars (A) are 

 more positive, and the bars (B) more negative, in the thermo- 

 electric series, than any other bar which can be formed of the 

 same metal. The thermo-electric force between the antimony 



(A) and the antimony (B), and between the bismuth (A) and 

 the bismuth (B), is pretty considerable. If a bar intermediate 

 between (A) and (B) be taken, that is to say, such that the direc- 

 tion of its length is otherwise inclined to the plane of most emi- 

 nent cleavage, or if it do not possess a regular crystalline tex- 

 ture, such a bar is negative with (A) and positive with (B) . 



This variability in the thermo-electric power of bismuth and 

 antimony seems to furnish a key to the explanation of the cur- 

 rents observed by Seebeck, Sturgeon and Matteucci, in circuits 

 formed of a single one of these metals. They have not been 

 explained hitherto. 



With regard to the direction of the currents between the warm 

 bismuth and the cold bismuth, the warm antimony and the cold 

 antimony, different experimenters have arrived at different results. 

 Vorsselmann de Heer, the last who has occupied himself with this 

 subject, has observed the current to pass sometimes from the 

 cold to the warm metal, and at other times from the warm metal 

 to the cold. He concluded from his observations, that the direc- 

 tion of the current depends on the greater or less difference of 

 temperature between the two bars. These cases of reversion 

 exhibited themselves in an especial manner with antimony. 



That such experiments should have any value, it is absolutely 

 necessary that the bars made use of should occupy the same 

 place in the thermo-electric series. Thus, for example, we must 

 compare (A) with (A), and (B) with (B), but not (A) with (B). 

 In the first place, it ought to be ascertained whether the two 

 bars be absolutely homogeneous. It is a remarkable fact, that 

 the deportment of (A) towards (A) is not the same as that of 



(B) towards (B). 



M. Svanberg's mode of experimenting was as follows : — The 

 two bars were fixed in copper handles, and these were connected 

 with a very sensitive galvanometer. Up to the point of contact 

 with the copper, the bars were enveloped in snow almost to the 

 free extremities. In this case, when the extremities are brought 

 into contact and then heated to any temperature whatever, there 

 ought to be no current; and this furnishes a test as to whether 

 the bars arc thcrmo-clcctrically homogeneous. But if before 

 bringing the bars into contact, the end of one of them be either 

 bested or cooled, a current is observed, the direction of which is 

 indicated by the galvanometer. If the two bars be of the bis- 

 muth (A) or of the antimony (A), the current proceeds from the 

 cold to the warm metal ; with the bars (B), however, the direc- 



