448 PROCEEDINGS OF THE AMERICAN ACADEMY. 



" I, a is not a constant but, in each of the liquids examined, a 

 function of both }-) and T. 



" II, In each liquid at constant temperature a increases with in- 

 crease of volume," that is, with decrease of pressure. 



Amagat's data do not, in the case of either liquid, serve for the 

 determination of a with certainty very near the boiling condition at 

 either high or low temperatures. On the other hand Professor Hall, 

 taking the data furnished by Regnault for the evaporation of ether 

 and of alcohol, and taking assumption 2 as holding through the process 

 of vaporization, found by means of a well-known formula the value 

 of a for this process at a number of temperatures in the case of each 

 liquid. The value of a obtained from any set of evaporation data 

 he called a\ and the following propositions were established with relation 

 to the values of a and a' : 



" III, In alcohol a is much less than a' at low temperatures, but 

 with rise of temperature the difference diminishes, a growing larger 

 and a' smaller. 



" IV, In ether «, at moderate pressures, is somewhat larger than a' ; 

 and both a and a' diminish slowly with rise of temperature, apparently 

 approaching equality." 



The formula by means of which a is calculated is 



a=gr-p),.^, 



in which e = the coefficient of expansion, 



^ = " " " compressibility, 



jT = " absolute temperature, 

 jo = " pressure on the liquid, 

 V =■ " specific volume of the liquid. 



The various quantities entering into this formula are expressed in 

 terms of the c. g. s. system. 



of temperature and pressure, we should expect both assumptions to hold. Per- 

 haps the natural interpretation of the fact that both cannot hold for ether or for 

 alcohol is the hypothesis that such groups do exist in each liquid, but that 

 their number is a function of both temperature and pressure. From this point 

 of view the magnitude of the variations which we find in a, when assumption 2 

 is held, maybe taken as a measure of the discordance of the two assumptions 

 and therefore some indication of the rate of variation of the number of groups 

 with variations of temperature and pressure. It might be better to keep assump- 

 tion 1, taking a as really constant, and see what change would be necessary in 

 tlie proposition given as assumption 2. 



It appears to be common opinion (see pp. 270 and 271 of the 1904 English 

 edition of Nernst's Theoretical Chemistr;/) that so-called polymerization exists in 

 liquid alcohol but not in liquid ether. E. H. H. 



