i8o SCIENCE PROGRESS. 



practical importance of the results of Raoult's work, which 

 he regarded as the most important aid furnished to the 

 chemist by physical methods since the time of Dulong and 

 Petit. In Meyer's laboratory Auwers devised a piece of 

 apparatus for conveniently determining the freezing-point 

 and made a series of observations. Other forms of 

 apparatus for use in the laboratory have been described by 

 Hollemann, Hentschel, Arrhenius, Beckmann, Eykman, 

 Fabinyi and Klobukow. Of these, the patterns due to 

 Eykman and Beckmann have found most acceptance ; in- 

 deed the form devised by the latter, its main feature being 

 a specially constructed thermometer, is now regarded as a 

 necessary part of the equipment of any well-appointed 

 laboratory. For very delicate measurements, Raoult, 

 Pickering, Loomis, H. C. Jones, and Lewis, have also 

 described pieces of apparatus. 



Attempts have been made to adapt the various modes 

 of estimating vapour-pressure to the determination of mole- 

 cular weights. In 1889 Beckmann tried to suitably modify 

 the statical method, but his work, as well as that of Tamman 

 and Loeb, clearly demonstrates the difficulties which stand 

 in the way of its adoption in the laboratory. Methods which 

 consist in estimating the weight of vapour given off from 

 solvent and solution have been used by Tamman, by Walker 

 at Ostwald's suggestion, by Will and Bredig, and by Beck- 

 mann. They can only be satisfactorily used with solutions 

 of medium concentration. Ostwald has proposed the em- 

 ployment of a hygrometer, and Charpy has made some 

 measurements with this instrument. By far the most con- 

 venient apparatus is based upon the dynamical or boiling- 

 point method of estimating vapour-pressure. Many of 

 Raoult's results were taken in this way, and in 1889 and 

 1 89 1 Beckmann described two handy pieces of apparatus 

 which admit of the determination of molecular weights from 

 the boiling-point of solutions with almost as much ease and 

 precision as from the freezing-point. Wylie and Sakurai 

 have also described similar methods. 



In what has been said regarding the boiling-point of 

 solutions it has been presupposed that, as compared with 



