346 PETER GUTHRIE TAIT 



temperature of the junctions. Was this law fulfilled through the range of 

 low temperatures which were the feature of Dewar and Fleming's experiments? 

 To settle this question Dickson first reduced the platinum thermometer 

 temperature readings to absolute scale, and then plotted with great care 

 on section paper the electromotive force of each metal lead couple against 

 the temperature of the junction whose temperature was varied. The curves 

 were parabolas with axis perpendicular to the temperature axis for gold, 

 silver, zinc and the alloy German silver, which metals therefore follow Tait's 

 law strictly. 



It was obvious at a glance that in the other cases the electromotive- 

 force curve was not a parabola with axis perpendicular to the temperature 

 axis. By graphical construction of the loci of the points of bisection of 

 sets of parallel chords, Dickson proved that these curves must be curves 

 of the second degree ; for the loci mentioned were very approximately 

 straight lines. In the case of the antimony-lead curve the loci of different 

 sets of parallel chords passed approximately through one point, indicating 

 that the curve was a hyperbola. In every other case the loci of mid-points 

 of different sets of parallel chords were parallel straight lines, indicating 

 that the curve was a parabola with axis parallel to these loci but not necessarily 

 perpendicular to the temperature axis. If following Dickson we designate 

 by the " Tait-line " the graph which gives the relation between the temperature 

 and the rate of change of the electromotive force per unit change of 

 temperature, the results obtained by him may be thus summarised : The 

 lead line being laid down horizontal, the Tait-lines for gold, silver, zinc 

 and German silver are straight, in accordance with Tait's theory ; for 

 platinum (two kinds, both of very great purity), copper, cadmium, nickel, 

 magnesium, palladium, and aluminium, the Tait-lines are cubical hyperbolas ; 

 and for antimony the Tait-line is a quartic curve. The two last statements 

 may be easily verified as properties of the parabola and hyperbola with axes 

 inclined to the temperature axis. 



In one of his early notes on thermoelectricity, embodied in his " First 

 Approximation to the Thermoelectric Diagram" (see Set. Pap., Vol. i, 

 p. 220), Tait remarked that 



When the temperatures were very high, the parabola was slightly steeper on 

 the hotter than on the colder side. This, however, was a deviation of very small 

 amount, and quite within the limits of error introduced by the altered resistance 

 of the circuit at the hotter parts, the deviations of the mercury thermometers from 



