206 Prof. H. L. Callendar on Platinum Thermometry. 



If, for instance, we take the observations of Fleming on very 

 pure iron between 0° and 200° C, and calculate a formula of 

 the Holborn and Wien, or Dickson, type to represent them, 

 we arrive at a curve similar to that shown in fig. 1 (p. 193). 

 (The values of the specific resistance of Fleming's wire are 

 reduced, for the sake of comparison, to the value R = 10,000 

 at 0° C.) This curve agrees very closely with that of Morris 

 and other observers between 0° and 200° C. The peculiarities 

 of the curve beyond this range are not due to errors in the 

 data, but to the unsuitable nature of the formula. A similar 

 result would be obtained in the case of iron by employing 

 any other sufficiently accurate data. It will be observed that 

 the formula leads to a maximum value of the temperature 

 £ = 334°, and makes the resistance vanish at —197°. Below 

 334° there are two values of the resistance for each value of 

 the temperature, and the value of dR/d£ at 334° is infinite, 

 both of which conditions are at present unknown in the case 

 of any metal, and are certainly not true in the case of 

 iron. If, instead of taking the value observed at + 196°*1 C, 

 we take the value obtained at the O.B.P. to calculate the 

 formula, we should find a better agreement with observation 

 at low temperatures, but the disagreement at higher tempe- 

 ratures would be greater. 



If, on the other hand, we take the same observations, namely, 

 c = '00625, and R/B,°= 2*372 when * = 196°'l, and calculate a 

 difference-formula of the type (2) corresponding to (3), we 

 find a 7 =-12-5, a = -005467, b = -000,007825. The points 

 marked in fig. 1 are calculated from this formula, and are 

 seen to be in practical agreement with the observations of 

 Morris up to 800°. As this formula stands the test of extra- 

 polation so much better than that of Holborn and Wien or 

 Dickson, we are justified in regarding it as being probably 

 more representative of the connexion between resistance and 

 temperature. 



Advantages of the Difference- Formula. — Mr. Dickson's ob- 

 jections to the platinum scale and to the difference-formula 

 appear to result from want of familiarity with the practical 

 use of the instrument. But as his remarks on this subject 

 are calculated to mislead others, it may be well to explain 

 briefly the advantages of the method, which was originally 

 devised with the object of saving the labour of reduction 

 involved in the use of ordinary empirical formulae, and of 

 rendering the results of observations with different instruments 

 directly and simply comparable. 



(1) In the first place, a properly constructed and adjusted 

 platinum thermometer reads directly in degrees of temperature 



