( 474 ) 



ration of tlie same plienoniena. Aftor the ituhlicatioii of Prof, van 

 DER Waals's paper I approached him prixalely with some objections 

 lo iiis views to wliieh he rejtlied in the most conrleous maimer. 

 Thoii,iiIi not completely satisfied it seemed nimecessai'v at the time 

 to pul>lisli my \iews until I shoidd he iji a j)ositioi) to a<ld lo (»iir 

 know ledu'e of the phenomena hy liirthci" ex|)eiMmenls. 



I have recently resumed the hivestigation and the resnlts obtained, 

 though naturally far from complete, seem of sufficient importance to 

 deserve an immediate pidtlication and to be discussed in conjiection 

 with previous experiments. 



For \arious reasons ^^■e had fixed our clioice on mixtures of hydro- 

 carbons — first of all ethane and ak-oiiols. llrielly oiii' results 

 WQVQ as follows: 



rHOH 



Fig 1. 



Foi' mixtures of ethane with ethyl-, propyl-, iso|)ro|»yl- and butyl- 

 alcohol there arc two temperatures A and /> (Fig. -1 ) be1\\ een which 

 three |)hases — two li(puds and vapour — are possible and the 

 critical (i.e. ])laitpoinl) curve accordingly consists of two bi-anches, 

 C\A and (\,B, (\ and (J^ representing the critical })oints of ethane 

 and alcohol r(^s|)ecfively and J /i the three-phase curve. Foi- ethyl- 

 alcohol .1 and /> are comparatively tar a[)art : for the highei- terms 

 of the series these points gradually approach each other ami with 

 aniylalcohol no separation into two li(pnds could be observed ; in 

 this case the critical curve was a continuous cui-ve Joining the two 

 critical points (_\ and ('., in the usual mannei'. 



For mixtures of methylalcohol and ethane we foun<l a branch (\A 



