286 Dr. R. A. Lehfeldt on the 



the alcohol, at any of the temperatures for which vapour- 

 pressure observations are recorded, so those data have to be 

 supplied from Alexejew's results *. Konowalow, in the 

 second part | of his paper, proceeds to show that the possible 

 forms of curve showing vapour-pressure against concentration 

 (temperature constant) are two : (i.) the flat part of the 

 curve bounded by a rising portion at one end and a falling 

 portion at the other ; (ii.) the flat part bounded by a falling 

 portion at each end. Isobutyl alcohol- water mixtures give a 

 curve of the latter kind. 



Isobutyl alcohol and water, however, possess a critical 

 point at about 130°, i. e., much above the boiling-point of 

 either. I therefore decided to study first mixtures of phenol 

 and water, which become homogeneous in any proportions 

 below 70°. 



The phenol was a commercial " pure " specimen ; to purify 

 it further, it was placed in a distillation-flask and melted ; 

 then air was drawn through it for about half an hour, whilst 

 its temperature was kept at about 160° to 170°, in order to 

 dry it. It w T as then distilled, and by far the larger part came 

 over between 178° and 180°. The fraction collected between 

 179 0, 5 and 180° (about half the mass) was used in the 

 experiments. To make up mixtures, the process always 

 adopted was to warm the stoppered bottle containing the 

 phenol to just above the melting-point, and pour the required 

 amount into a wei o-hing-flask. It was found that the moisture 

 absorbed from the air during the process was quite inappre- 

 ciable. The phenol, kept day after day at 40° to 50° ready 

 for use, slowly turned pink, showing the presence of rosolic 

 acid ; but a comparative colour-observation showed that the 

 amount of impurity was probablv not more than 1/10,000. 

 When it was necessary to estimate phenol in a mixture, that 

 was done by the method of Koppeschaar, tribromphenol 

 being formed and the excess of bromine replaced by iodine 

 and titrated with thiosulphate. The method gave quite 

 satisfactory results. 



Experimental Methods used. 



The measurements on phenol-mixtures gave results con- 

 trary to my expectations, so that I became suspicious of the 

 experimental methods. In the end I made use of four 

 different kinds of apparatus, but found that they gave results 

 in practical agreement, so that it became chiefly a question of 

 convenience to decide between them. 



* Wied. Ann. xxviii. p. 315. f Wied. Ann. xiv. p. 222. 



