Properties of Liquid Mixtures. 287 



The first method tried was the " dynamic," carried out 

 with the same apparatus as described in Part II. It required 

 no modification, except the use of a new thermometer, since 

 the old one did not go above 60°. The new thermometer was 

 a longer one, graduated in ^ from 0° to 100° (by C. E. 

 Midler, ^No. 8). Its corrections were obtained in two ways : 

 first by comparison with a standard (Reichsanstalt, 7347) at 

 certain fixed temperatures, viz., the boiling-points of methyl 

 acetate (57°), methyl alcohol (65°), and ethyl alcohol (78°); 

 secondly, by measurements of the vapour-pressures of water 

 under the same conditions as in the actual experiments ; in 

 these conditions part of the stem w r as exposed. 



To use the apparatus the required mixture was weighed out 

 from melted phenol and distilled water, then warmed up in 

 the w T eighing-bottle until it became homogeneous, and poured 

 into the tube of the vapour-pressure apparatus. The apparatus 

 works satisfactorily except for mixtures on a very steep part 

 of the curve of vapour-pressure (p) over concentration (z) ; 

 w T hen dpjdz is great, the change of composition of the liquid, 

 due to the evaporation becomes disproportionately important, 

 and the static method is to be preferred ; in the case of phenol 

 mixtures, however, that only affects a small part of the range 

 — mixtures with 90 per cent, or more of phenol. 



The anomalous result that made me at first doubtful of the 

 accuracy of the method was that up to 60 or 70 per cent, of 

 phenol added to water made practically no difference to the 

 vapour-pressure of the water. To check this, I made one or 

 two experiments by the static method, in a barometer tube 

 surrounded by an alcohol-vapour jacket of the usual pattern. 

 They were not carried out with any attempt at accuracy, but 

 sufficed to show that the previous observations could not be 

 far wrong. The problem then was to determine the small 

 difference in pressure between water and the phenol- water 

 mixtures, and as for that purpose a differential gauge is 

 obviously more appropriate, I set about designing and making 

 the apparatus described below. Its design is based on a point 

 of technique that does not seem to be much known, and to 

 which, therefore, I should like to draw attention. If a glass 

 tube be drawn out fine, sealed at one end, and evacuated, the 

 sealed end may be broken under the surface of a liquid, 

 which then flows in at any desired rate according to the 

 diameter of the tube, and the tube may at any moment be 

 fused off in the middle by a mouth blowpipe, without any 

 inconvenience whatever. This process of filling with a liquid 

 will I think be found advantageous in many cases. The only 

 trouble about it is to get the canillary of the right bore"; 



X2 



