160 Proceedings of Indiana Academy of Science 



actly balanced by the return into the liquid phase of the gaseous mole- 

 cules. In the event that the system under consideration is not closed 

 and that the rate of escape of the liquid molecules over-balances the rate 

 of return of the gasecus molecules, the equilibrium otherwise prevailing 

 is disturbed and the liquid evaporates. 



The rate of the process of evaporation is then subject to several 

 variable factors. In the case of liquid vi'ater, for example, exposed in 

 an open vessel to the atmosphere, the rate of evaporation is directly 

 proportional to the velocity of air currents, to the radius of the ex- 

 posed surface, and to the vapor tension of the liquid. The latter is in 

 turn directly proportional to the tempei'ature. Further, the rate is 

 inversely proportional to the distance of the surface of the liquid from 

 the rim of the containing vessel, to the relative humidity or partial 

 pressure of aqueous vapor of the atmosphere, and to the barometric pres- 

 sure. Various "laws of evaporation" have been proposed in the effort to 

 connect these variables with the rate of evaporation. Some of the ex- 

 pressions developed include such constants as the molecular weight of 

 the liquid and the latent heat of vaporization, if the rate is to be stated 

 in terms of mass of liquid evaporated per unit time. 



In the case of solutions there must be considered also the variation 

 of the vapor pressure of the solvent with the amount of the dissolved 

 solute. It is well known that the vapor pressure of a solvent is lowered 

 by the addition of a soluble salt, for example, and that the retardation 

 in the rate of evaporation thereby produced is proportional to the 

 amount of salt dissolved. Furthermore, it has been shown recently 

 that loss of solvent by evaporation of equimolar solutions is least with 

 salts that give the greatest number of ions. 



Experimental Procedure. The burettes used in this work were of 

 the type recommended by the Bureau of Standards, and were sealed 

 off at the bottom in order to prevent any error from leakage. Two sets 

 were arranged, the individual members being chosen as nearly as pos- 

 sible with the same inside diameters. Over the "open" set was sus- 

 pended a cover about 10 cm. from the ends of the burettes in order to 

 protect them from falling particles of dust. Over the ends of the 

 "covered" set were placed ends of test tubes, 6 cm. long, to serve as caps. 



The temperature at the time of reading the burettes was taken from 

 a theiTnometer suspended among them. A sling psychrometer was used 

 for determining the relative humidity of the atmosphere according to 

 the method of the U. S. Weather Bureau. The barometric readings 

 are uncorrected. 



Working under the conditions of experimentation stated, determi- 

 nations have been made of the effect upon the rate of evaporation of 

 (a) the size of the burette, (b) the concentration of solutions, and (c) 

 the presence of caps over the ends of the bui'ettes. For the study of 

 (a) several liquids having distinctly different vapor tensions were 

 selected — water, benzene and toluene; for (b) a series of standard solu- 

 tions of sodium chloride in water were prepared by dissolving the proper 

 weight of salt; and for (c) parallel series were run for (a) and (b) — 

 one set covered and the other open. 



