XXXVl INTRODUCTION. 



Table 35. Pressure of aqueous vapor — English measures. 



This table is a conversion into English measures of Table 36. It gives 

 the vapor pressure in saturated air for temperatures varying b}^ of2 from 

 — 20^0 to 2i4.°o Fahrenheit. 



The tabular values are given in inches to four decimals. 



A column of differences for o.°i is added for convenience in interpolating. 



Tables 36, 43. Pressiwe of aqueous vapor. — Metric tneasures. 



These tables, taken from Broch, give the pressure of aqueous vapor to 

 hundredths of a millimetre for temperatures var>dng b}^ o.°i C. from — 29°o 

 to ioo.°9 Centigrade. The values for temperatures between 0° C. and 45° C. 

 are given in Table 43, the remainder in Table 36. 



Table 37. Pressure of aqueous vapor at low temperature. — (C F. Marvin.^ 



Broch's vapor pressures at temperatures below 0° C. (32° /^) as given in 

 Tables 35 and 36, when compared with the actual observed values of Reg- 

 nault are found to be systematically too large. This discrepancy signifies 

 that the etnpirical formula adopted by Broch fails to represent accurately 

 the law of variation of vapor pressure for temperatures both above and 

 below the freezing point. Moreo\»£r, the failure in the application of the 

 formula might be inferred from the laws of diffusion following from the 

 kinetic theory of gases, for these give no reason to suppose that the function 

 expressing the relation between vapor pressure and temperature is continuous 

 between the two states of water and ice. 



Under proper conditions water can be cooled far below 0° C. (32° F.) 

 before solidifying,, so that at the same temperature we may have it either 

 in the liquid or the solid state, and experiments confirm the theory of 

 diffusion in showing that the pressure of the vapor is different according as 

 it is in contact with its liquid or its solid at the same temperature. The 

 method hitherto employed of combining vapor pressures above and below 

 freezing, and attempting to represent them by a single continuous function, 

 must therefore be considered as radically erroneous. 



Recognizing the systematic errors of the vapor pressures given by 

 Broch's formula for temperature below freezing, the Chief Signal Officer 

 lately authorized a new determination by direct observation. This experi- 

 mental investigation has been carried out by Prof. C. F. Marvin, from the 

 results of which {Anmcal Report Chief Signal Officer, 1891; Appendix No. 

 10,) Table 37 is reproduced. The interpolation between the observed press- 

 ures which were noted at intervals of about 5° F., was effected graphically 

 and not by mathematical formula. 



The vapor pressures were determined for the case of the vapor in con- 

 tact with ice and not a water surface. For the temperature of melting ice 

 (0° C. or 32° /\) all values agree. Below this temperature Marvin's vapor 

 pressures are slightly smaller than Regnault's, but differ from the latter less 

 than any other tabular values. 



