760 



MR J. D. HAMILTON DICKSON ON 



first and last observations, and thus gave the greatest possible range. The mid-points 



of these chords are shown on fig. 10, below the short line a, to which they are parallel. 



Their co-ordinates are given in the annexed Table VII. From these the equation of the 



nearest straight line passing through them was calculated by least squares and found to 



be- 

 ll +2-7201T=674-352 ..... (41) 



Table VIL 



T. 



H. 



T. 



H. 



154-5 



252-5 



168-0 



217-5 



1560 



249-5 



170-0 



212-5 



158-0 



244-5 



171-5 



207-5 



159-5 



241-0 



1730 



203-5 



161-5 



236-0 



175-5 



197-0 



1633 



231-5 



178-5 



190-5 



1650 



226-3 



1803 



182-3 



166-3 



221-0 



181-8 



179-3 



Hence o>, the angle made by the axis of the curve with the axis of E.M.F. is given by 



cot w = 2-7201, whence <* =20° 11' 6" .... (42) 



The sum of the sixteen errors obtained by comparing the calculated and observed values 

 of H was + 5'9 - 6'0, thus verifying the accuracy of the calculation; and the probable 

 error of an observed value of H was ±'64. 



Having obtained the value of tan «, the calculation of the equation of the parabola 

 followed the same course as in the case of Platinum (Callendar's). Fourteen observations 

 were employed out of the thirty recorded. The equation found was 



= 528-4516 i (H - 79-40325) - T '^^ \ ■ • ■ • (43) 



in which, if we replace H and T by their values in terms of E and t, it takes the form, 



more suitable for calculation, 



E= 194656-49 + 544-020- 21975-42^0 .... (44) 

 where 0=170-71226-!! ..... (45) 



Hence the equation of the Tait-line is 



dE K .. no 10987 71 

 _=- 544-02 + — r - 



(46) 



The sum of the errors between the calculated and observed values of H was + 1188- 

 1170, which, being practically zero, verified the accuracy of the calculation. From 

 these errors the probable error was ±2*84 micro-volts over a range of some 750 microvolts. 

 The Tait-line is like that of platinum, and is a cubical hyperbola ; it is shown on fig. 

 10. The portion connected with these experiments is the continuous line AB. Tlie 



