2 3 o SCIENCE PROGRESS. 



the molecule, which in turn depends on the distance at which 

 it can attract its neighbours. - 



Inasmuch as Ramsay and Shields' analogy cannot be 

 justified until these points have been disposed of, it seems to 

 me that, even on neglecting other reasons, such as discrep- 

 ancies between the testimony of viscosity and surface energy, 

 no stress should be put upon the actual values of liquid mole- 

 cular weights as deduced by the method at present employed. 



Vapour - pressure and boiling-point. — The vapour- 

 pressures of the fatty acids have been determined by 

 Kahlbaum (20), who finds that the statical and dynamical 

 methods give the same results. Landolt's statical obser- 

 vations, which have long been quoted as supporting the 

 opposite conclusion, are shown to be erroneous, probably 

 because his liquids were neither quite dry nor air-free. 



Batelli's statement that at constant temperature the observed 

 value of the vapour-pressure depends on the relative 

 volumes of liquid and vapour present in the experimental 

 tube has been tested by Young (21) in the case of iso- 

 pentane. Over wide variations in the ratio of the volumes 

 there was no indication whatever of this effect. 



Wirkner (22) has measured the vapour-pressures of a 

 number of aromatic compounds. In several cases the 

 curves cut one another, and he finds that none of the usual 

 vapour-pressure formulae satisfactorily represents the facts. 

 A new and complex formula, involving the heat of vapor- 

 isation and the specific heats of the vapour, has been 

 devised by Krsevitch (23), but it is little suited for general use. 



A matter of practical importance, which is also treated 

 theoretically, has been taken up by Barell, Young, and 

 Thomas (24), who deal with the separation of three liquids 

 by fractional distillation. Applying F. D. Brown's views to 

 the case of three liquids, after making assumptions regarding 

 the constants in the formulae, they find that if a mixture of 

 three liquids, A, B, C, having increasing boiling-points, be 

 fractionated, the lower fractions soon consist of A and B, and 

 the higher of B and C. These may be fractionated in the 

 usual way. This result is verified in the case of a mixture 

 of methyl, ethyl, and propyl acetates. 



