Measurements of Precision in Platinum Thermometry. 545 



When Professor Mather and the author made experiments 

 with the Ayrton-Jones current balance *, it was found that 

 some fine-stranded flexible leads (quite new) were not 

 constant in their resistance because of the comparatively 

 large number of places at which the strands were broken. 

 The difficulty was overcome by replacing the leads by others 

 of less flexibility and containing strands of copper 038 mm. 

 in diameter. 



As a practical way of eliminating the changes in the 

 relative resistance of thermometer leads, we suggest the 

 abolition of leads of fine strands of wire, and the substitution 

 of leads, preferably short, each of one conducting wire of 

 copper. The leads should be strengthened near the ends so as 

 to fix strong terminals to them, and the four leads might be 

 bound together so as to ensure all being at practically the 

 same mean temperature. 



In putting forward this suggestion we are aware that tl e 

 importance of keeping the compensating leads low in 

 resistance and as invariable as possible has not been over- 

 looked by all workers in platinum thermometry, but never- 

 theless it is still the common practice, even in precision 

 work, to use flexible leads of the type we have experimented 

 with. The abolition of such leads cannot fail to lead to 

 greater accuracy in temperature measurements and to add 

 to the already great utility of the Callendar-Grifnths bridge. 



The Heating Effect of the Current. 



Passing now to the heating effect of the measuring 

 current. That the resistance of a platinum thermometer is 

 increased by an appreciable amount when the usual measuring 

 current is passed through it was recognized in the earliest 

 days of platinum thermometry. In 1887 Profes.-or Calendar f 

 estimated that his thermometers increased in temperature by 

 o, 01b' C. when a current of O'Ol ampere passed through 

 them. In 1899 % Dr. Chree pointed out that the heating 

 effect due to a given number of watts expended in the wire 

 was not the same at all temperatures, there being a diminu- 

 tion in the heating effect as the temperature rises. It was 

 necessarv to point this out as some observers at that time 

 proposed to keep the watts constant at all temperatures. 



In 1902 § Prof. Callendar made a very complete statement 

 on the heating effect. In his paper on " Continuous Elec- 

 trical Calorimetry " he states " the cooling effect of conduction 



* Ayrton, Mather & Smith, Phil. Trans. A. ccxxiv. p. 523 (1908). 



t Callendar, Phil. Trans. A. clxxviii. p. 184 (1887). 



X Chree, Proc. Rov. Soe. vol. lxvii. (1900). 



§ Callendar, Phil. Trans. A. cxcix. p. 55 (1902). 



