( 335 ) 



the vapour pressure with the temperature. Flic atmospheric pressure 

 on the mercury in I lie open manometer was read on an aneroid 

 barometer which we occasionally compared with a mercury baro- 

 meter. 



§ 3. The determination of temperature <m<l the degree of accuracy. 



For measuring the temperature we have used for the determina- 

 tions in the neighbourhood of the boiling point of oxygen the 

 resistance thermometer Pt' j which has been compared l ) (comp. 

 Comms. N°. 95' Sept. '06 and N°. 101" Dec. '07) over a large range 

 with the hydrogen thermometer of Comm. N°. 95* (Oct. '06) which 

 we shall call Hill ; tor --182° this was done on March 25 th 1907 

 (comp. Comm. N°. K)\ n table I). Some calibrations have afterwards 

 been made, which together with those just named are given in table II. 

 This table also contains the results of an indirect comparison of 

 Pt'i with another hydrogen thermometer, which we call T 4 and 

 which was used in an investigation of Dr. Fkustel and one of us 

 (K. 0.) with a differential thermometer helium-hydrogen, which 

 investigation will soon be published. 



In connection with § 4 of Comm. N°. 10i a we have also given 

 in table I the results of 4 comparisons between two resistance 

 thermometers Pi' i and PH' d 3 ) at about —183° C. and —217° C. 



For the measurements on IS 1 ' and 19 th Dec. '07 B m , Pt' 'j and 



i) W P ti= 1.01806 Wpt'l (comp. § 2 of Comm. N°. 101« Dec. 1907). 



-) This resistance thermometer is the same as the one which in Comm. N". 99& 

 (Sept. '07) is called Pt'd. Afterwards it lias broken in the middle, and consequently 

 the resistance was diminished to half its value (Pt'd). Here it is indicated with 

 two accents because it has then undergone a small reparation owing to which 

 Pt"d= 1.00073 Pt'd at 0°. 



8 ) This value has been reduced with the factor of the preceding footnote. 



