﻿306 Mr. Baly and Dr. Ramsay on the Relations of Pressure, 



Aniagat, in 1883 {Annates de Chimie et Physique, xxviii. 

 p. 480), undertook experiments to test the results obtained by 

 Siljestrom and Mendeleeff, but was unable to confirm them. 

 His conclusion is, that for the lowest pressures measured 

 (6*54 millim.) the differences were sometimes positive, some- 

 times negative ; and that the necessary error of experiment 

 was of an order of magnitude equal to that of the differences 

 found. One of the main difficulties of observation was found 

 in the diffraction of the walls of the manometer in which he 

 measured the pressures to which the gases were exposed. 

 He appears, however, to have been successful in overcoming 

 this difficulty, by the use of polished plane surfaces ; but it is 

 noteworthy that a suspicion of this source of error does not 

 appear to have occurred to either of the former experimenters. 

 Aniagat also states that the utmost limit of measurement is 

 probably overestimated at O'Ol millim. ; and indeed it is 

 very difficult, if not impossible, to judge of the contact of two 

 lines, the cross-wire of the cathetometer and the meniscus of 

 the mercury in the gauge, to within such a minute distance. 

 One of his criticisms of MendeleefFs paper is worth quoting: — 



" In M. Mendeleeff s experiments, a set of values of pv are 

 obtained corresponding to a tolerably limited range of pres- 

 sures ; this would seem favourable to a knowledge of the 

 direction in which these products alter ; but it is necessary to 

 guard against possible illusion in this matter. 



" A constant cause of error may affect all these products ; 

 for as they deal with smaller and smaller values the effect 

 will be more and more marked, and a regular variation may 

 be conceived to exist, with no foundation in fact. Such an 

 effect would be produced, for example, if the vacuum in the 

 barometer were not absolute. It is certain, that whatever 

 precautions be taken, this vacuum is not absolute ; the effect 

 of this is to diminish the real value of the pressures, that is to 

 say, too small pressures are read : hence the value of pjp' is 

 reduced, and since p is smaller than p\ the values of pv/p'v' 

 are rendered too small ; and as the error increases with 

 reduction of pressure, they will ultimately be less than unity." 



Amagat proceeds to point out that even the vapour-pressure 

 of mercury ceases to be negligible when a few hundredths 

 of a millimetre cause the supposed deviation. 



In 1886, C. Bohr (Wied. Ann. xxvii. p. 459) measured the 

 pressures and volumes of rarefied oxygen. His conclusions are 

 that at a temperature of ll c, 4 oxygen deviates from Boyle's 

 law. For values of pressure greater than 0*7 millim., 

 (p + 0*109) v — k ; and for low r er pressures, (p + 0'070) v = k. 

 At 0*7 millim. therefore oxygen undergoes a sudden change 



