This paper is cle\otcd to the dircct d e t e r m i~l~~t i o n of the \apor prcisures of solutions of the nitrates of silver, of a m r n o~i i~~m , and of lithiu~ll, a t temperatures varying from 30" C. to 105' C. ancl a t conce~itrations \.ar)i~ig from 10 to 85 weight % (for 11thulm nitrate, the limited solubilit) preclucled measi1reil1ent> beyolid (5%) From the vapor prcsci~res, the eiithalpiei of evaporat~on of water (by a ~l~o d i f~r a t i o n of the Clapeprotl-Clau-ills e q~~a t i o n ) , thc difterential herits of dilution, and the activ~ties of \later (.dration; a t extremel~. high concentration there will be strong competition for the remaining \vatel-molecules. Using the concept of Stokes ancl Robinson (8) that the ions aclsorb la,-ers of niater molecules in accorclance with the Brunauer, E m n~e t ,ancl Teller equation, it should become progressively inore difficult to strip away the successive la>-ers of water surrouncling the ions; the last aiicl most fir~iilj-bound la>.er should req~lire a large amount of energy to remove it. Application of the Clapeyron-Clausius equation (though not ill its normal forin: see Discussion) to the vapor pressures of concentrated -.solutions should give tlie heat of evaporation a~i d hence, perhaps, a ine:~sui-e of the difficulty of removing water from the ions.The problem of finding vapor pressures \i7as first attempted I ) ! . fincling boiling points a t different fixed pressures. After sl~eiicling a year, ho\vever, on this method of attack we abandoned it because we fouiicl that, a l t h o u~h -we could control pressure and boiling point to an!. desired degree of constancy, we hat1 110 certainty as to the coiiiposition of the equilibl-i~iiii liquid. Indeed, the work about to be describecl slionred that we occasioiiall~. had composition errors ecl~~ivaleiit to an error of 0.5" in boiling point. --Because of this uncertainty in the conipositioii variable of the boiling point method, we cleciclecl to turn to a method in which the composition factor \vould be known and invariable. The direct determinatioii of vapor pressure in a closed system oflers such a methocl. The loss of water by evaporation to produce one atmosphere pressure in a closecl s)stenI is negligible, even in the ...