14 Messrs. C. W. Waidner and F. Mallory on 



For the platinum thermometer used throughout this in- 

 vestigation a change in temperature of o, 001 C. corresponds 

 to a change in resistance of about O'OOOOl ohm. Owing to 

 the great sensibility of the galvanometer, changes in re- 

 sistance far more minute than the vernier would indicate 

 (*. e. 0"00001 ohm) were easily detected. Advantage was 

 taken of the high sensibility of the galvanometer to reduce 

 the current used for measuring the resistance of the coil of the 

 platinum thermometer to 0*002 or 0'003 of an ampere, thus 

 almost entirely eliminating any heating up of the coil by the 

 measuring current, which was only kept on for a few seconds 

 at most. Perhaps an equally satisfactory method of reducing 

 to a minimum the uncertain effects of heating of the platinum 

 coil by the measuring current would be to so regulate the 

 current through the coil that the energy used up in heating 

 the coil is always the same at all temperatures. 



Constants of Platinum Thermometer. 



Determination of B (the resistance of the platinum ther- 

 mometer at 0° C). — The ice used in these determinations 

 was that manufactured by the Diamond Ice Co., of Baltimore, 

 and was very clear and pure. The water used in this ice is 

 first filtered ; it is then frozen in large forms (long and high, 

 but of shallow depth) from the sides towards the centre. 

 When a thin sheet of water remains at the centre, the 

 freezing is stopped and this water drawn off, thus insuring 

 ice of great purity provided it is kept from contact with the 

 freezing-mixture. Conductivity tests made on water obtained 

 from this ice show a high specific resistance. 



The thermometer was inserted in a mixture of this ice 

 (pounded very fine) and distilled water. In the first and 

 second determinations of B a double-wall vessel was used, 

 the inner vessel being nickel-plated. Ice and water were 

 placed in both compartments. The outer vessel in all the 

 zero determinations of the platinum, as well as for the mer- 

 curial thermometers, was wrapped with heavy boiler-felt to 

 a thickness of about 4 cms. 



The second and third determinations (Tables V. and VI.) 

 were made in the same mixture of ice and distilled water, the 

 only difference being that in the third the inner vessel was 

 removed. The higher resistance obtained in this case (corre- 

 sponding to about o- 002 C.) strongly suggests that, in a 

 room whose temperature is 20° C. above that of the ther- 

 mometer, the effect of radiation cannot be neglected. We 

 were at first of the opinion that the removal of the inner 



