1909-10.] The Vapour Pressures of Mercury. 
521 
XXXVII. — The Vapour Pressures of Mercury. By Alexander 
Smith and Alan W. C. Menzies. 
(MS. received June 6, 1910. Read July 4, 1910.) 
An examination of the existing data of the vapour pressures of mercury 
shows that at the higher temperatures a most astonishing lack of precision 
characterises all the measurements. The existing tables, especially above 
300°, are therefore completely untrustworthy, a fact which might be 
guessed from the disagreement between the values they give (see table 
below). Regnault made his measurements with the thermometer bulb 
completely immersed in 50 kilos of the violently “ bumping” metal. 
Hence his results are much too high. Ramsay and Young’s well-known 
values were based on observations at four points. The two lowest of 
these were original : one was the boiling-point, taken from discordant data 
of Regnault’s ; and one was a measurement of the pressure at the boiling- 
point of sulphur. For the last, Regnault’s temperature was taken, and 
Callendar and Griffiths have shown that the value is 4° too high — a change 
which alters the vapour pressure of mercury at 445° by about 140 mm. 
Young (J. Chem. Soc., 59 (1891), 629) has modified the table to take 
account of Callendar and Griffiths’ work. But he assumes that an un- 
screened bulb in an unjacketed bath of sulphur vapour has the same 
temperature as was attained by Callendar and Griffiths with most careful 
jacketing. His temperature at 445° is therefore still about 2° too high. 
Gebhardt (Berichte deutsch. physik. Gesell., 7 (1905), 184), with no mention 
of precautions or corrections in respect to temperature or pressure readings, 
publishes data which, in view of our results, are probably on the whole 
nearly 1° too low. The average deviation of a single observation from a 
smooth curve is 1‘2°. Cailletet, Colardeau, and Riviere ( Comptes Rendus , 
130 (1900), 1585) measure the vapour pressures from 400° up to 880°, 
without stating what values they assumed for the fundamental points in 
their temperature scale. 
In the present work the isoteniscope, platinum resistance thermometer 
and gauge described in the preceding paper were employed. Forty-three 
measurements between 250° and 435° were made. An accuracy of ±0*1° 
was aimed at. The results are represented by the following Kirchoff- 
Rankine formula : — 
log ^ = 9-9073436 -3276-628/0 -0-6519904 log 0. 
