f 4.SI ) 



of olliaiH": IIk' associ.-ilioii Ikiwcnci- Ik'coiiics urndiKilh sii-oiiüci' : llic 

 <li|i in llic siirlacc (•.•iiiscd liv lliis'i l;i-;i(Iii;iI1\ iiKulilics llic ^liapc (tf 

 llic cross-itlail : llic |»l;ul|M>iiil passes lliroiiuii a iiiiiiiiiiiiiii pi-cssiirc 

 l)('l\V('(Mi 25' and 'M)\ and die crilical curve dien Wcuins lo i-isc 

 ra|)i(M_\ . Tlic end ot' die cross-plai! lliiis clianLics w idioiil a discond- 

 nidlv inin a li(pd(l-plail : in llic ni(>an dnic die main plail u'ocs (Hi 

 d('\ ('lopinu' on llic appi'(»a(di of llic cridcal poini of cdianc: as cxplainc»! 

 a sniad sid)sidiai-y plail is lornicd \\ Idcli app(\-\i"s al .1 on tlic |n-acl- 

 ical pari ol' llie sniTacc. l*i*ol>altly aii cxcliaiiuc of' plailpoinis occurs 

 oji die dicoi-clical pari. of llic same iialure as w itli die luLiiier alco- 

 hols, die resiill iieinii' dial al low lenipei'alui-es die cross-plail is ciil 

 lliroiiLili Ity oiié seU'-conlainod Ii(pnd-plail. IJiil as far as llie plieii- 

 (Hueiia ai-e concerned lliis is (Miliroly ininialorial. 



As far as die (Wperimenls could l>e carried (i. e. iip lo 275 

 alniosplieirs) llie ci-ilical ciirv(^ condnuod lo rise towards die lel'l, so 

 lli;il llioi-e is no indicalion ot' llie cxisleiice of a dilferenl |)lailpoinl. 

 riic ra|)idily willi which llio inixiii,<i-i)rcssiire intM'oases is Iridy 

 remarkahlc. 



If we compare the ii<>iires i\w iiielliylalcohol and for Ihe iiiuhei- 

 lernis, a cerlaiii i-esenil»lance w ill he noticed, especiady if we do nol 

 assume the contraction of the closed |ilail to nolhinu in the latter 

 case. The association tends to produce the same moditicalion in 

 the usual diaii'ram in itoth cases, hut the actiiiLi" causes appear lo he 

 much nioK^ eifeclixc with methylalcohol - the sti-ojiuer association 

 of the alcohol, [)ossihly a smallei- mutual allraction or the influence 

 suuu'esled h\ I'l'of. V AN DKiMV A Ai.s of the small molecular \-oliiiiie of 

 the alcohol may conlrihule to this result. Kor lids suhslance the 

 |)laitpoijit remains outside the cross-plait at low ItMnperaliires, wiili 

 tli<^ others it succeeds in disappearinu' inside. AN'hethei- inside this 

 plait any chanu-es take place similai- to those occurriiiLi- w illi inelhyl- 

 alc(tliol on the practical part of the surfac(> we cannot tell. Ihil in 

 any case I liaxc assured myself that with elhyIalcoh(tl a new plait- 

 point cur\e does not a|>pear down to 7<S : ethylalcohol and ethane 

 remain miscihie in all pro|)ortions. 



Mt'tlii/hilc(ili(il <in(l clJuitii' ini.r hi/ inrssiirc. in this res|)ect they 

 «•onlrast with mixtures phenol and water for which the licpiid plail 

 ahovc a certain tempei-ature far helow die critical region separates 

 c<tinplelely from the cross-plait and thus has a plaitpoini on Ihe side 

 of the |tosili\-e /--axis. In view of the prohahle disappearance of the 

 association at liiiili lemperaliires it is possihie that in the lallei' case 



1) Gontinuitüt etc. II. p. I'Jl. 



