226 SCIENCE PROGRESS. 



dioxide as determined by Chappuis (13), who employs a 

 new method based on measurements of the refractive index, 

 by means of interference bands, is 31*4° ; Amagat found 

 3 1 "35°. That two totally different methods should give 

 practically identical values is surely strong evidence that 

 the critical temperature is a function of the nature of the 

 substance alone. The most weighty experimental testi- 

 mony is thus in favour of the generally accepted view that 

 at the critical point the substance is in a definite uniform 

 condition, and that above the critical temperature one — 

 and only one — value of the volume corresponds with a 

 definite value of the pressure and temperature, although 

 De Heen especially is still active in seeking evidence to 

 the contrary (8). 



LIQUIDS. 



Molecular complexity. — On taking- a survey of the 

 physical chemistry of liquids for the past year, the most 

 noticeable feature is the general application of the idea that 

 liquids differ in respect of the complexity of their funda- 

 mental molecules. Whether liquefaction is attended by 

 molecular aggregation has long been a puzzle to chemist 

 and to physicist alike. To Eotvos, however, belongs the 

 credit of first attempting to get a definite answer to the 

 question, and it is by means of Eotvos's method that 

 Ramsay and Shields, and Ramsay and Aston (14) have 

 recently served to direct attention to this subject. Accord- 

 ing to the method employed it is assumed that if K, the 

 temperature coefficient of the molecular surface energy of 

 a liquid, be 2*1 21, the liquid consists of simple molecules; 

 that if K be less, the liquid consists of molecular aggregates, 

 and that the degree of complexity can be estimated from the 

 value of K. In this way it is found that the great majority 

 of liquids consist of simple molecules, but in other cases the 

 liquid molecule is more complex than the gaseous molecule, 

 the remarkable point being that these associated liquids, 

 with one or two trifling exceptions, are hydroxy compounds. 

 Here are the formulae which are supposed to express the 

 average molecular weight of some of the associated liquids 



