( 333 ) 



by T l and T 7 , . First of all on account of the theoretical existence of 



hidden three-phase- pressure; but also because the presence or absence 



of a hidden plaitpoinl does no! coincide with the intersection or 



dhp </'■> ,, „ 



non-intersection of - = and = 0. .Moreover the existence ol 



da? <!<- 



the liquid sialc is also assumed here for all temperatures, however 

 low. The occurrence of the solid state may, of course, he a hindrance 

 for the observation of what we have called the temperature 1\ . 

 Thus in mixtures of water and phenol an upmost temperature limii 

 has been found for three-phase-pressure; but it is still an open 

 question whether there also exists a lowest temperature, higher than 

 the absolute zero point. 



( To In 1 continued 



Physics. — "On the measurement of very low temperatures. XXI. 

 On the standardizing of temperatures by means of boiling 

 points of pure substances. The determination of the vapour 

 pressure of oxygen at three temperatures." By Dr. II. Kamer- 

 linou Onnes and Dr. C. Braak. Communication N°. 107° from 

 the physical laboratory at Leiden. 



(Communicated in the meeting of May 3o, 190S). 



§ 1. Introduction. In a preceding Communication N°. J 01" (Dec. '07 j 

 we have spoken of the desirability of determining once for all certain 

 temperatures by means of boiling point apparatus, because the points 

 of the temperature scale thus fixed have the advantage over those 

 fixed with resistance thermometers and thermo-elements that they 

 do not depend on the durability of special apparatus and they facilitate 

 comparisons between thermometers in different laboratories. 



This Communication treats of a number of determinations with 

 oxygen: <i. a little above and a little below the normal boiling point 

 from which the latter could be derived; l>. at .'*<>() and 51(5 mm. 

 mercury pressure which may serve to give information about the 

 further course of the vapour pressure curve. 



After some preliminary determination-- we have constructed two 

 apparatus of different dimensions, in each of which different quan- 

 tities of gas could be successively condensed. Thus we have obtained 

 two series of independent determinations; ai the same time the 

 purity of the gas could be tested. 



