On the Density of Fluid Bismuth. 295 



P : Q = R : S. This of course is only applicable when none of 

 the coils used have any appreciable self-induction. 



Another method is to reverse the battery-current. 



Let P t Q T be the values of P Q when the current is in one 

 direction, P 2 Q 2 when it is in the other. Then the true value 



ot i is i(^+H 



A third method, applicable to coils with no self-induction, is to 

 interchange the battery and the galvanometer. The thermo- 

 electromotive force is then superposed on that of the battery, 

 and the condition PS = QR holds. The most serious objection 

 to this seems to be that the whole current must pass through 

 the sliding junction, which may thus get considerably heated, 

 and so damage the bridge-wire. Perhaps the simplest cor- 

 rection is to keep the battery resistance low compared with 

 the others. 



I tried the effect of using copper wire for the bridge in 

 order to have only one metal throughout ; but I found that 

 the contact between my fingers of two pieces of copper wire 

 of somewhat different thickness produced a very appreciable 

 current. 



In the experiments given above the E.M.F. produced is 

 about equal to that of a copper and platinum couple in which 

 the difference of temperature is between 2° and 3°. 



I was assisted in making the experiments by Mr. E. B. 

 Sargant, Scholar of Trinity College. 



XL. Determination of the Density of Fluid Bismuth by means 

 of the Oncosimeter. By Professor W. Chandler Roberts, 

 F.R.S., and Thomas Wrightson, Memb. Inst. C.JE* 



SOME time since, one of us described the results of experi- 

 ments made to determine the density of metallic silver 

 and of certain alloys of silver and copper when in a molten 

 state "I". The method adopted was that devised by Mr. E. 

 Mallett ; and the details were as follows : — A conical vessel of 

 best thin Low-Moor plate (1 millim. thick), about 16 centims. 

 in height, and having an internal volume of about 540 cubic 

 centims., was weighed, first empty, and subsequently when 



* Communicated by the Physical Society, having been read at the 

 Meeting 1 on February 12th. 



t Roberts, Proc. Roy. Soc. vol. xxiii. p. 493. 

 % Proc. Roy. (Soc. vol. xxii. p. 306, and vol. xxiii. p. 209, 



Y2 



