as determined ~by Five Platinum Resistance Thermometers. 219 



lying one under the other in a vertical plane beneath the grass of the 

 south laAvn of the Radcliffe Observatory, and within a few feet of the 

 Stevenson's screen in which the dry bulb and the wet bulb, the maxi- 

 mum and minimum thermometers, are suspended. A fifth thermometer 

 was subsequently placed at a depth of about 10 feet in a separate pit. 

 The actual depths of the various thermometers as measured in October, 

 1898, were as follows : 



Thermometer. I. II. III. IV. V. 



Depth 6| iii. 1 ft. 6 in. 3ft. 6 in. 5 ft. 8| in. 9 ft. 11^ in. 



The resistance box is in its general design similar to that described 

 by Mr. Griffiths,* but simplified to suit the particular class of work 

 for which it was intended. It is provided with three principal coils, 

 A, B, and C, whose nominal values are, 20, 40, and 80 box units 

 respectively, a box unit being about 0*01 ohm. 



The apparatus is provided with a slow motion contact maker, of 

 Mr. Horace Darwin's pattern,! and Mr. Griffiths's thermo-electric key 4 



In the standardisation of the apparatus the method described by 

 Mr. Griffiths, in his article in ' Nature,' referred to above, was in the 

 main followed. The temperature coefficient was determined by 

 Mr. Griffiths, in his own laboratory at Cambridge. Two separate 

 series of observations led to the following results : 



Range of Temperature 



Date. temperature. coefficient. 



July 27 9-18 0-000242 



Augusts 12-51 0-000240 



The value actually used in the reductions was 0-00024. 



From observations made when the instrument was mounted in situ 

 at Oxford, the values of the coils were found to be 



C = 80-1581 



B = 39-979 S-mean box units, 



A - 19-863J 



and one scale division of the bridge wire is equal to 

 1-0134 mean box units. 



One of the most important considerations in connection with this 

 subject is the degree of permanence in the fundamental points, as 

 determined at considerable intervals of time; but the process of 

 standardisation is not one that can be very frequently applied. 



* ' Nature,' rol. 53, November 14, 1895. 

 t ' Nature,' vol. 53, November 14, 1895. 

 t ' Phil. Trans.,' A, pp. 397-8, vol. 184 (1893). 



