1919-20.] Measurement of Dissociation Pressures. 159 
7 80 mm. The only two non-preliminary measurements by the tensimetric 
method gave 7*72 and 7*76 mm. : average, 7‘74 mm. In spite of precautions 
to avoid condensation * of saturated water vapour before reaching the 
absorption tubes, the gas current saturation results would be expected 
to turn out high by a fraction of 1 per cent., in conformity with the 
consistent results of the extremely careful work of Berkeley and Hartley.] - 
As the results stand, therefore, it is obvious that the discrepancy is well 
within the experimental error of such measurements. And thus the 
anomaly disappears. 
In the interest of progress, it is perhaps allowable briefly to indicate 
one or two points at which error may have entered into the experimental 
work of previous investigators in this field. Too much confidence has 
perhaps been placed upon the tensimetric results of Frowein, which fall 
over 5 per cent, below those given above. A comparison of such other 
tensimetric measurements as have been published since Frowein’s time 
points in the same direction, thus indicating, in the writer’s opinion, that 
Frowein too lightly assumed the absence of permanent gas from his tensi- 
meters. The ingenious suggestion of Campbell, based on the fact that the 
observed vapour pressure of water in air is less than in a vacuum, thus 
overshoots the mark in harmonising the results of Tammann with those of 
Frowein, for Frowein’s figures are unquestionably low. Schottky seems to 
have understood the danger of the presence of permanent gas, for he allowed 
his tensimeters to lie in the horizontal position, with the opposite sides 
in free communication, for twelve hours prior to erection for observation. 
When, after erection, one bulb of the tensimeter is raised to the higher 
temperature, as was done in his work, not only will the dissociation 
pressure on that side increase, but so also will the air pressure. But air is 
soluble in paraffin oil, which Schottky used, as well as in all common 
manometric liquids except mercury. One would therefore predict, in 
accordance with Henry’s law, that Schottky should find, exactly what he 
did find, an initial excess of pressure that slowly fell off to the true 
equilibrium value as the air pressures on opposite sides slowly became 
equalised by the mechanism of solution and diffusion of air in the oil. 
Others also, up to the present time, continue to fail to recognise this 
consequence of Henry’s law.J With regard to Partington’s work, it is 
suggested that glass-wool plugs § should not be used to filter rather small 
volumes of air of 100 per cent, humidity, if the humidity is later to be 
* Gf. Menzies, J. Amer. Chem. Soc ., xlii, p. 978 (1920). 
f Trans. Roy. Soc. London , A, ccix, p. 177 (1909). 
J The matter will be referred to elsewhere. § Gf. Menzies, loc. cit. 
