198 IOWA ACADEMY OF SCIENCE 



the use of oil. The drying continued for three days and nights 

 at a temperature of at least 200°. with a Gaede mercury pump 

 continually running, and in the presence of phosphorus pen- 

 toxide. 



The data for the curve, figure 20, connecting equilibrium 

 pressure and total mass of water, were taken over a period of 

 seven or eight months, and while the curve is not yet complete 

 to saturation the experiment has been carried far enough to re- 

 veal several matters of interest, mentioned in the summary. In 

 this curve the separate feeds are numbered. Following any one 

 feed more readings than one were of course taken, sometimes a 

 series of readings extending over considerable time. 



SUMMARY OF RESULTS. 



1. The curve is of particular interest because of the larger 

 number of separate feeds before the half saturation region is 

 reached, showing a refinement much greater than in the Trouton 

 method. 



2. The curve is divided into three distinct parts : A, B, C-D. 



3. An examination of the ratio y/x 2 for part A of the curve 

 shows that this part approaches very close to the parabolic form. 



4. Part B of the curve may be taken as linear, and when pro- 

 jected it passes through the origin. 



5. Part C-D is a region of more or less instability. 



6. In part A of the curve more time was required, in general, 

 for the pressue to reach the equilibrium value than in the other 

 region, where the equilibrium pressure ensued in a relatively 

 short time after a feed. 



7. Although the pressure data used in the curve were cor- 

 rected to a constant temperature of 22° C. there still appears a 

 marked fluctuation of pressure with the actual temperature pre- 

 vailing at the time the reading was taken, and generally in the 

 direction one would expect. This effect not only appears in the 

 main curve shown, but also in auxiliary curves using series of 

 readings taken between actual feeds. The effect is to we ex- 

 plained by the change with temperature of the mass of water 

 present as vapor, while the pressure correction for temperature 

 assumed a constant mass of water vapor. There would appear 

 to be sufficient data for information as to how the mass of water 

 in the condensed film varies with temperature. 



