752 



MR J. D. HAMILTON DICKSON ON 



This point being settled, the equation of the parabola was calculated from sixteen 

 observations of the thirty-one recorded and gave 



(t + 305-20283) 2 = -5631819(166252-47 - E) .... (25) 



Hence we get the equation of the Tait-line, namely, 



2(^ + 305-20283) 



dE 



dt '' 



■5631819 



(26) 



The sum of the errors was 4-1125 — 1123, which, being practically nil, confirms the 



accuracy of the calculation. Hence the probable error was found to be I "56 micro-volts 



over a range of 2670 micro- volts, a notable result. 



Equations (25) and (26) may be expressed more conveniently for calculation in the 



forms 



E= 166252-47 -1-775625^ ..... (27) 

 where 



and 



or, more usually, 



(28) 



= t + 305-20283, ... 

 dE/dt = -3-55125(9 

 dE/dt= -3-55125*- 1083-85 ..... (29) 



From equation (25) the B-parabola was drawn, and along it the whole of the 

 observations are marked by means of small crosses. From the A-curve the values 

 of dE/dt in Table IV. were measured by means of the thread. They were determined 

 with great care, each temperature being read by the help of four deviates of 2, 3, 4, 

 and 5 mm. ; they are plotted by means of small circles along the Tait-line (straight) of 

 equation (20). 



Table IV. 



t'C. 



100 dE/dt. 



t'C. 



100 dE/dt. 



74| 



168-1 



- 801 



99-1 



36 



151-1 



-1011 



89-2 



H 



137-8 



-148f 



67-3 



- 2H 



125-0 



-172| 



58-9 



- 51$ 



113-6 







The vertex of the parabola falls within the limits of the diagram, namely at 

 t= -305-2°, E = 166252-47. 



Copper. 



On account of the purity of copper it was one of the first metals examined. In 

 the earliest attempts the nature of the E.M.F. curves was sought for from the form 

 of the Tait-line, obtained from measurements made on the hand-drawn A-curves ; 

 and if the test indicated a parabolic curve, an attempt was made to determine the 

 focus and directrix. Such an investigation for copper gave the Tait-line as almost 



