2* Mr. T. Andrews. Electromotive Force from [May 8, 



Thomson effect in any given metal are dependent mainly, if not 

 entirely, upon the specific heat, specific resistance, and coefficient of 

 expansion of the metal (or upon changes of these properties with 

 changes of temperature), the Thomson coefficient is not exactly 

 given by the expression spec, heat X spec. res. x 10' (Exp.-f-34) 2 , 

 even if due allowance is made for the uncertainty of the numbers in 

 columns II, III, IV, and VI, and for the fact that some of them may 

 vary greatly with different specimens of the same metal. But I have 

 not succeeded in finding a better expression.* 



II. "Experimental Research on the Electromotive Force from 

 Difference of Potential during Diffusion in Tidal Streams." 

 By THOMAS ANDREWS, F.R.S.E., Assoc. M. Inst C.E., 

 Wortley Iron Works, near Sheffield. Communicated by 

 Professor STOKES, Sec. R.S. Received April 20, 1884. 



An examination of the composition of the waters throughout the 

 length of a tidal stream during diffusion of salt and fresh water 

 consequent on tidal action, reveals a very considerable difference 

 in the proportion of saline constituents between the water at the 

 surface and that at the bottom, during certain times of tide this 

 difference amounting sometimes to near 100 per cent., and it may 

 frequently be either much greater or less according to tidal fluctua- 

 tions. 



This fact constitutes the basis of the investigation which the author 

 undertook, to obtain some approximate quantitative measurement of 

 the resultant electromotive force, &c., arising from such difference of 

 potential. It is known that a current is set up when a bar or plate of 

 the same metal is immersed in two dissimilar solutions in contact, 

 one capable of acting readily upon the metal, the other having little 

 or no action on it, the whole forming a circuit. 



The current continues until diffusion renders the composition of 

 the solutions uniform, after which a reverse current may not 



With regard to aluminium, which in column X comes between platinum and 

 magnesium, instead of between lead and tin, as in column II, it is possible that the 

 specific resistance given by Matthiessen as 0'029 is too high. I do not know of any 

 other determination professing to be exact, but it is stated in Wurtz's "Die. de 

 Chimie," p. 129, upon the authority of Deville, to be one-eighth that of iron, which 

 would make it 0'012. This, however, is undoubtedly too low. If it were 0-019, the 

 place of aluminium in column X would be the same as in column II, and if it were 

 as high as 0'026, its place would be between magnesium and lead. I may mention 

 that some experiments of my own show that the coefficient of the Thomson effect in 

 aluminium comes slightly above that of lead, instead of below it. And in the 

 diagram at p. 178 of Jenkin's " Electricity " the inclination of the aluminium line 

 is also shown as positive. 



