THERMO-ELECTRIC DIAGRAM FROM - 200° C. TO 100° C. 765 



where 



= ^ + 853-437 ..... (51) 



Hence the equation of the Tait-line is 



dE 1A000 6 270215-07 , 4 



w -- 10,883-3 + —J0- ■ - (52) 



The sum of the errors between the observed and the calculated values was + 2073 — 2071 , 

 practically verifying the calculation. From these the probable error was 4 - 45 micro- 

 volts. Considering the unusually great range of the observations for this metal, namely, 

 5900 micro-volts, the theory may be taken to agree with the observations. 



The B-curves, in fig. 12, are the calculated sheared curve, the E.M.F. parabola, and 

 the Tait-line. Along the two former curves the observations are marked by small 

 crosses. 



Magnesium. 



On plotting the observations for magnesium it was found necessary to "shear" the 

 curve, in order to determine its nature, and the slope of its axis if it should turn out 

 to be a parabola. For this purpose the co-ordinates chosen were 



T = < + 200, 



H = |j + 2T- 300 = ^ + 2^ + 100. 



The observations were found to be very erratic between 10° and 100° C, so much so 

 that it was difficult to make anything definite out of them. An attempt to draw the 

 A-curve of E.M.F. was made, and as a further help the sheared curve also was drawn, 

 thus affording two views of the deviations of the observations, and enabling the two 

 curves to be fitted into correspondence as fairly as possible. These two A-curves are 

 shown on fig. 13. From the sheared A-curve the mid-points of a series of seventeen 

 chords parallel to the axis of T were carefully observed, the highest chord being that 

 for which H = 230. These points are shown on the diagram below the short line a to 

 which they are parallel. To make allowance for the uncertainty of the upper end of 

 the sheared curve, the mid-points of the chords forH = 210, 220, 230, were not used to 

 determine the line of mid-points ; the remaining fourteen mid-points for H = 70, . . . 

 200, were alone employed, and by least squares, gave the equation 



H-8-5643,74T= -1171-2203 .... (53) 



A second series of mid-points of parallel chords was observed ; these also are shown 

 on the diagram below the short line b, to which they are parallel, and which makes an 

 angle with the axis of T whose tangent is — \. They also lay closely on a straight 

 line to the left of those marked under a ; these two rows of points being clearly 

 parallel, allow us to consider the curve to be a parabola. In accordance with 

 equation (53), the axis of the sheared parabola makes an angle « with the axis of 

 E.M.F. such that cot a> = 8*564374 ; this value was adopted in calculating the equation 

 of the parabola. 



