CONDEXSATIOX AND CRITICAL PHENOMENA. 261 



substances and threat retardation in the establishment of the 

 equlHbrium in the long and narrow experimental tubes. If 

 the mixture is partly liquid and is then compressed to a 

 smaller volume, the pressure will be higher than it should 

 be, and in leaving the mixture for some time at the same 

 volume the quantity of the liquid increases while the pres- 

 sure diminishes. If we start from a smaller volume and 

 increase the volume the pressure will be too low and rises 

 slowly while some more liquid evaporates. It takes a very 

 long- time for these changes to come to an end and it is 

 very difficult without waiting hours at every change of 

 volume to obtain concordant values at diminishing: and 

 increasing volume. Similar uncertainty attaches to the 

 determination of the points d and e and of the critical 

 constants. If the mixture is not homogeneous the critical 

 temperature may be found several degrees too high. 



Similar results were obtained by Ramsay and Young 

 working with a mixture of alcohol and ether. Their in- 

 vestigation brings out the influence of retardation and 

 imperfect mixing very strongly. 



Some isothermals for mixtures of carbonic acid and 

 sulphurous acid were obtained by Bliimcke, but his experi- 

 ments do not go as far as the critical region. Ansdell 

 studied the condensation and critical point of mixtures of 

 carbonic acid and hydrochloric acid. A number of observa- 

 tions regarding mixtures, especially critical temperatures, 

 were made by Dewar. Unfortunately, several of the results 

 obtained are doubtful in consequence of the influence of 

 imperfect mixing of the substances. 



Critical points of mixtures of liquids were determined 

 by Strauss, Pawlewski, G. C. Schmidt, and others. Differ- 

 ences between mixtures and pure substances were not 

 noticed : in fact the method of observation was, as a rule, 

 one which did not allow either of a measurement of the 

 pressure or of a change of volume. 



The above remarks show in what way the experimental 

 method had to be improved in order to get definite results 

 independent of any retardation in the establishment of 

 equilibrium. The little iron stirring-rod (if necessary, en- 



