( 629 ) 



solutions tlie diininulioii oC llic vapour lonsioii w itii iuci'oaso of con- 

 centration increased more than tlic coucentratiou, whereas for tlie 

 nitrates of K, Na and for KCiO., the reverse was ol)served. 'I'iie 

 Later investigations of 1>i{kmkk^), IIkt.miioltz'^), Wai.kkr') and Dn^'/rKKiei ') 

 gave the same result as those of their predecessors viz. this that in 

 genei'al the diniiuution of the \apour tension iu.'i'cascs nioi-e rapidly 

 than the concentration, and thai llic salts which form an cxccplion 

 to this ride are chiefly the anhydi'ous nitrates. 



rp to HMKi the determinations of the fi-eezinii, |toint of salt 

 solutions of smaller coiicenli-ations than those which had Ixmmi earlier 

 investigated, yielded the ivsidl, that the molecidar lowering of the 

 freezi]ig point dcciuutscs with ijici-ease of the concentration. As these 

 measurements wei'e contimied to the concentration of ± 1. gi'. mol. j)er 

 1000 gr. H./), it followed necessarily, that wdiere the reverse course 

 had been ascertained by the earlier observers, a miinmum xalue had 

 to occur in the molecular decrease of the freezing [)oint, l)nt foi- a 

 concentration which lay above that where the investigators of later 

 time had stopped. 



In i89(^, li()we\er, 1 had come to the conclusion in the delernu- 

 nations of the va[)onr tension by means of the micromanometer ^), 

 that foi- dilute solutio]is, i.e. for solutions below the concentration 

 of 1 gr. mol. |»er iOOO gr. II.^O, the mol(M'idar dinnmition of the 

 vapour tension iiu-nui.scd with the concenti'atiou. This was found 

 inter alia for solutions of NaCl, KOH, H^SO^ and (InSO^, whereas 

 for KN( ).,-solntions the reversed course was found. This result was 

 therefore in jjerfect accordance with what had been found by my 

 predecessors for more concentrated solutions, but was dii-ectly o|)posed 

 (except KNO3) to the results of the determinations of the freezing 

 point. 



The (piestion was now: '"Which rc'sulls are the correct (Uies?" 



In the determinatio]! of the boiling point 1 hoped to lind a means 

 to answer this question. After lia\i)ig ap|»lied s(niie im|»ro\-euienls 

 to the method and aftei' haxing rendered myself indepcMident of the 

 fluctuations of the atmos|)heric |U'essni'e by using a iikihosIhI, Wwix-aw 

 the investigations and the result was published in April 1900"). 



1) Rec. tr. Ghim. 6, 122. (1887). 



2) Wied. Ann. 27, 568. (1886). 



3) Zeitschr. f. phys. Chem. 2, 602. (1888). 



*) Wied. Ann. 4^,513.(1893); ib. 62. 616. (1807\ Ann. phys. Chem. 27, 4. (1808). 



5) Archiv. Néerl. (2) 1 (18U7). 



'■■) These Proceedings U April 21 11)00 p. 635, April 20 I'lOl, 111 p. 717. In lliu 

 same year (1900) Jones, Ch.vmbkus and Fr-.v/KU cryoscopically found minima for 

 MgCls and BaCl^ lying at 0.1—0.2 gr. mol. 



