72 Mr. J. Parker on Thermoelectric Phenomena. 



Any sensitive galvanometer whose time of vibration exceeded 

 two seconds, and in which there was not excessive damping, 

 would probably" be suitable for this determination ; if the 

 period is too small, among other difficulties, that of self-in- 

 duction, delaying the passage of the permanent current, would 

 occur. 



With the apparatus connected as shown to a galvanometer 

 whose resistance is known, the actual determination need not 

 take more than two or three minutes, and involves but very 

 little more trouble than is necessary to determine the capacity 

 in terms of a standard condenser, without reposing the confi- 

 dence which is required in that case. With a galvanometer 

 not specially chosen for the purpose, the capacity of a half 

 microfarad condenser, tested on several different occasions by 

 different observers, gave "497 for the lowest value and '501 

 for the highest ; but tested by the ordinary steady-deflexion 

 method, the value determined was only '485, the error in 

 this case being due entirely to the imperfect elasticity of the 

 short fibre suspension. 



The plan is not, however, suggested as a means of stand- 

 ardizing condensers where more refined means are available, 

 but in the numerous cases where long-fibre ballistic galvano- 

 meters are not at hand and where an accuracy of, say, one 

 half or one quarter per cent, is deemed sufficient. 



IX. On Thermoelectric Phenomena. By J. Parkee, B.A., 

 late Scholar of St. John's College, Cambridge^. 



THE present short paper is necessary to complete the paper 

 which I published on the same subject in the Philoso- 

 phical Magazine in October 1888. 



In that paper, a constant k was discovered which was tacitly 

 supposed to have different values for different metals. Since 

 then, however, owing to a suggestion made by Prof. Liveing, 

 I have been led to reconsider the subject, and I now find that 

 the constant k must have the same value for all metals. This 

 remarkable discovery introduces some important simplifi- 

 cations. 



Let two condenser-plates, x, ?/, of any the same metal be 

 joined by three wires, A, B, A', as in the figure ; A, A' being 

 of the same metal as ,x and ?/, and B being different. Let the 

 plates (x, ?/) and the junction of the wires (B, A') be kept at 

 the same absolute temperature tQ, and the junction of the 

 wires (A, B) at the absolute temperature t, where t — to is 

 indefinitely small, and = t, say. 



* Communicated by the Author. 



