178 



Mr. H. E. Schmitz. On the Determination of [Feb. 2, 



and the writer has pleasure in expressing his obligations to Professor 

 'Schuster for advice during the progress of the work. 



Platinum TJiermometer. 



The temperature of the liquid air was obtained by means of a 

 'Calendar's platinum thermometer with exposed coil, made in accord- 

 ance with instructions kindly given to the author by Mr. J. E. Petavel. 

 It was found that, when the instrument was placed in a narrow test- 

 tube containing the last remains of a quantity of liquid air made some 

 time before, its resistance went up to a maximum at which it remained 

 practically constant until nearly all the liquid air had evaporated. 

 The steady maximum obtained on several occasions in this way was 

 taken as the resistance at - 182'5, which (disregarding the variation 

 with pressure) is very nearly the temperature of liquid oxygen. This 

 value, together with the resistances at and 100, enabled Callendar's 

 formula to be used. The lowest observed temperature of the liquid 

 air obtained in this way was - 191 0> 0. 



Calorimeter for Method of Mixtures. 



The inner vessel was of very thin brass or copper. Different 

 -quantities of water were used, as shown in the following list : 



In each case the vessel was nearly full of water, and the thermometer 

 was immersed to a constant depth. The water equivalent of the 

 thermometer bulb (from the mass of water displaced) was 0-9 gramme; 

 the water equivalent of the whole immersed portion was estimated at 

 T5 grammes. 



Mercury Tliermometer. 



This was one of Goetze's Beckmann thermometers. It was carefully 



compared with a secondary standard of which the calibration cor- 



tjons were known; in the final comparison the readings were taken 



at every tenth of a degree in a very slowly rising temperature. As a 



