216 Profs. Ramsay and Young on the 



the results obtained by these two observers are contradictory. 

 Zambiasi found that the temperature at which the meniscus 

 disappears on heating is identical with that at which mist and 

 strise form on cooling; and that both temperatures are lowered 

 by increasing the quantity of the liquid relatively to the 

 vapour. De Heen, on the contrary, found that the tempera- 

 ture at which the meniscus disappears is raised on increasing 

 the relative quantity of liquid ; while Galitzine found the 

 temperature at which the meniscus disappears to be higher 

 than the temperature at which it appears, but both of them 

 practically independent of the relative quantity of liquid. 



Again, the critical temperatures of the esters given by 

 de Heen differ enormously from those observed hy Nadejdine 

 and by one of us ; that of methyl acetate differs by no less 

 than 46° I 



There can be little doubt that in most, if not in all of such 

 experiments, the arrangements employed to maintain constant 

 high temperatures must have been sadly defective. Battelli 

 made use of fractions of petroleum-oil, boiling at intervals 

 of temperature of 10°. This plan was tried by one of us as 

 long ago as 1879, and subsequently rejected as entirely 

 unsatisfactory. The heating arrangement employed by Galit- 

 zine would be satisfactory if the sample of naphthalene used 

 as a jacket were pure. This method is a modification of that 

 which we have constantly employed, and frequently recom- 

 mended, during the last ten years ; yet it is strange that while 

 Galitzine refers to an early paper by one of us containing 

 conclusions which have been acknowledged as erroneous in 

 subsequent papers, he makes no reference to our later work, 

 of which he appears to be ignorant. Moreover he relies on 

 de Heen's work, which is very inaccurate. Precisely such 

 conclusions as appear in these papers were given by one of us 

 in the paper referred to (Proc. Roy. Soc. 1880, xxx. p. 323, 

 and xxxi. p. 194), in which similar arguments were adduced 

 in favour of liquid molecules, and of the non-identity of 

 the liquid and gaseous states at . and above the critical 

 point. The author of that paper again takes this opportunity 

 to say that he no longer believes in the conclusions ; and 

 he would regret the publication of that paper were it not 

 for the hope that it may serve as a beacon to warn others, 

 beginning to experiment on the subject, from the numerous 

 rocks and shoals with which it is beset. 



Since that paper was published we have been engaged on 

 many pieces of joint work, and the critical constants of ether 

 and of the alcohols have been determined with great care ; 

 the constants for no less than 18 other liquids have also been 



