54 Mr. H. H. Poole on the Thermal Conductivity 



this way generally only differs sliglith from what would be 

 obtained by neglecting a. 



The thermoelectric properties of the wire- used for the 

 couples were tested at the National Physical Laboratory, 

 and from the figures thence obtained a cnrve was carefully 

 plotted showing the B.M.F. of a single couple whose cold 

 junction is at 0° C. and whose hot one is at 0° 0. A- in 

 practice the cold junction temperature is that of tin- key 0\ 

 an easily calculated correction hastobeadded to the observed 

 E.M.F. Ex to obtain the value of the E.M.F. E 2 corre- 

 sponding to a cold junction at zero, from which the tempe- 

 rature 6 may be read off to the nearest degree on reference 

 to the curve. If the distance of the single junction from the 

 centre were the geometric mean of the distances of the inner 

 and outer junctions of the multiplecouple, the temperature 

 of the single junction would he the mean of the temperatures 

 of the other junctions. This condition is sufficiently nearly 

 realized. A second curve is plotted showing the thermo- 

 electric power at various temperatures. From this, knowing 

 the mean temperature of the junctions of the multiple couple 

 we obtain the difference of temperature A indicated by a given 

 E.M.F. E 3 . 



In obtaining a the total change of temperature i- often 

 very small, so the increase in E 2 is divided by the thermo- 

 electric power for the given temperature and by the time in 

 minutes between the two sets of observations. The time was 

 always read at the moment of reading Ej. 



The following table (No. I.) gives a typical day's work. 

 The first column gives the time, the second the position of 

 the indicator on the core of the choking-coil, and the third 

 the reading of the wattmeter measuring the energy supplied 

 to the furnace. The power in watt- is about tour limes the 

 reading of the wattmeter. The other columns -how the 

 values of quantities already mentioned : under K i- found 

 the conductivity deduced from a single set of readings, v hile 

 under K' is that obtained From a pair of sets correcting for 

 temperature change. In certain cases where several hours 

 have elapsed between successive reading- ami where the 

 temperature change in that time is -mall, the value obtained 

 for K from the la-t single Bet el' reading- may be nearer the 

 truth than the value K'. as towards the end the temperature 

 change must be very -low except for irregular changes doe 

 to voltage variation. For this reason in adopting a mean 

 result lor K lor a day's work, due regard is paid to the figures 

 in both columns K and \\ . 



