However, the percent error of the BPE determinations, hence 
osmotic coefficient determination, is larger at lower 
salinities. Hence, measurements below 0.3% salinities 
would be much less accurate. 
The BPE data obtained in this work have been compared 
with some of the recently published data on boiling point 
elevations of sea water solutions. Figure 21 shows a 
comparison of the-various data, typically at 100°C. <The 
relative values are the same at the other temperatures. 
Stoughton and Lietzke (13), used osmotic coefficients 
of synthetic sea salt solutions measured at 25°C to evaluate 
the constants of their empirical extension of the Debye- 
Huckel equation. Their values come nearest to the observed 
values, within 0.5 to 1.0%, deviating more at the higher 
concentrations. Their empirical expression seems, to, bexa 
good approximation to the actual values, though not 
accurate to £0.001°C. 
Fabuss and Korosi (14) assumed that the relative molal 
vapor pressure depressions of the solutes can be added to 
give the total contribution for the vapor pressure depression 
of the solution. From this they computed the boiling point 
elevations. At higher concentrations, their assumption is 
not valid, as seen £rom Figure Zl. Moreover, they report 
BPE(S tO) an accuracy of = 0701; G only. 
Grunberg (15) computes BPE's from vapor pressure 
depression data obtained experimentally on sea water. He 
used "Ca free'' sea water, which was kept in a bath whose 
temperature was maintained to only +0.01°C. These values 
are corrected for "Cav to get values for natural, sea water: 
He also reports BPE data to +0.01°C. The deviations at the 
higher concentrations are rather large. 
Lindsay and Liu (16,17) very recently measured the vapor 
pressure of simulated sea water in concentration ranges 
above 1.5 normal sea water. They have used their experi- 
mental values to fit a double polynomial in salinity and 
temperature to the BPE data obtained. A comparison of the 
actual data obtained by them with the data obtained in 
present work is given in Table 28. There is a large 
variation between the data of Lindsay and Liu and this work. 
One of the reasons might be the replacement of Ca*t+ ions by 
Mgt+ ions. Much data has been taken by Lindsay and Liu in 
comparatively higher temperature and concentration ranges 
under which conditions CaSO4 would have precipitated out of 
66 
