260 Professor Fleming [June 5, 



thermo-electric power of the metals with respect to each other at 

 that temperature. If we measure the slope of the electromotive 

 force curve at any point, it can easily be shown that the numeri- 

 cal value of this slope gives us the rate of change of electro- 

 motive force with temperature. If we plot these slopes in terms of 

 the corresponding temperature, we obtain another set of curves called 

 curves of thermo-electric power. Lead is always taken as the stan- 

 dard metal for comparison, because the Thomson effect in lead is zero. 

 From our chart of thermo-electromotive forces we have constructed 

 another one of thermo-electric powers (see Fig. 14). The lines of 

 thermo-electric power cut the lead line in various places, and the 

 temperature at which they do this is called the neutral temperature 

 of that metal with respect to lead. Professor Tait deduced from his 

 experiments that these thermo-electric lines were straight lines for 

 temperatures above zero Centigrade, and he made, in addition, the 

 important discovery that for certain metals such as iron and nickel 

 the thermo-electric lines have sudden changes of direction at high 

 temperatures. 



The general result of our investigations at low temperatures is to 

 show that, whilst in some cases the thermo-electric lines, as may be 

 seen from the diagram in Fig. 14, are approximately straight lines 

 for temperatures down to the lowest reached, they are not all by 

 any means straight lines. In some cases, such as iron and bismuth, 

 we find sudden changes of direction of the thermo-electric lines 

 similar to those found by Professor Tait at higher temperatures, 

 and this indicates a change in sign in the Thomson effect at that 

 point. Moreover, in many cases there is a decided tendency of the 

 lines of many metals to bend round in a manner which indicates 

 that their thermo-electric power probably would become zero at the 

 absolute zero of temperature. 



The temperature at which the thermo-electric line of any metal 

 crosses the line of lead gives us the neutral temperature of that metal 

 with respect to lead, and at that temperature the metal is thermo- 

 electrically identical with lead. If one junction of a couple is at a 

 temperature as far above the neutral temperature of the metals as 

 the other is below it, the couple will give no electromotive force. 

 This provides us with an experimental method of determining the 

 position of certain neutral points. Thus, for instance, if one 

 junction of a platinum-zinc couple is placed in liquid air and the 

 other is raised to above 30° we get no electromotive force from that 

 couple. This indicates that the neutral temperature of platinum and 

 zinc is about — 85'', and this is shown to be the case from the chart. 

 Two general conclusions are arrived at from a study of the thermo- 

 electric lines as laid down in our chart. The first of these is that 

 the thermo-electric lines of many metals are by no means straight 

 lines over extreme ranges of temperature. Hence the thermo-electric 

 power is not simply a linear function of the absolute temperature. 

 The second important fact is, that in the thermo-electric lines of 



