( 574 ) 



at n (Icliiiilo placo hf^eonies denser and (icnser'i. Al. a U'iven inomeiil 

 "sli'iae" appear, the suhstaiice boils and rains, a nieniscMis is Conned. 

 Fi'om the \(M'v first, tlic l.-ittci- is pertectlv sharp and Hat, it remains 

 at the sani(^ ph-ice also after stirrinp,'. The tein])ei'atni'(' was ohsei'X'ed 

 iinnicMliatelv after tlie appearance of the striae, then the pressni-e and 

 the volnmc ixwd then the tcinperatnre was detei'iniiie<l auain. So we 

 ohtain th(> pressni'e for one or a i'vw hnndredths of a degree below 

 the critical 1('in])eratiir(\ Kor this a eori'eclion has Ixmmi applied. 



If A\(' allow the leinpcratnrc to rise, we slill see a nieniscns 

 dnrinu' a considerable lenuth of time after the tcmpcratnre in the 

 bath has risen to a few hnndredths of a deui-ee altox-e the ci'ilical 

 tcmperalnre. It disa|)pears immediatelv, howexcr. when we s|ii'. 

 Hence this is no phenomenon of eqnilibriiim. 



So ^^<' coidd dei-i\-e the lemperatiii'e and the |)i-essnre <lirectlv 

 from obserx jitioji. Kor the xolnme this \\ as not possible, as owing 

 to the ellect of graxitation the phenomena described above occnr 

 at different xobimes. so for instance in mv observations with a 

 volnme : ().()()ol>24 (slii-i-er below), where the meinscns appeared 

 in the immediate neighbonrhood of the top of the Inbe, and with a 

 N'olnme ().()()4'2S[ (stii-rer alxtxei where Hie nieniscns ap|>eared 1 mm. 

 above the niercnrv. 



The method genei-ally nseil foi' the determination of the critical 

 volnme consists in determining some li(pnd and vaponi- densities 

 at teni])eratnres below the ci-itical. and then nsing the rnle of the 

 rectilinear diameter of ('aii.i.ktkt and Mathias. To this end we 

 have drawn on a diagi-am the densities i-esnlting from \\ic data of 

 § 4. table XXI and § 2. r, a- a t'nnction of the lemperatni-e, and 

 the diameter has been draxxji on it. A <le^■ialion from tlie rectiiinearity 

 conld not be stated w illi certainty. Kor the critical density, exjti'essed 

 in terms of the theoretical noi-mal deirsiiy we <leri\e from this diagram 

 2:39, hence for th(> ci-itical volnme 0. 00418. 



Another method is this: At the critical pond ,^„ =: .^,-, 



(for the proof see for instam-e ('omm. N". 75). 



'i\> determine - we have combined in table XXII nnder (f 



the \apoin- |)ressnres resnlting fi'om s^ 2 : foi' the temjteratnres 25^.55 C' 

 and 2<S .15 (' we haxc talcen the means b-om tlu^ different A'alnes 

 for the pressnre. 



1) If the temperature is kc])! eonstaiit, the luisl dnrinti' a consideralile length 

 of lime (say 10 minnles) docs jiol eliaiis^e to llie eye. 



