Page images
PDF
EPUB

shows how the sensitiveness for a given change of capacity decreases as we approach the position of resonance.

Figs. 3, 4, 5, and 6 were now drawn, in which the abscissæ represent temperatures, and the ordinates the capacities of

[merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]

the water-condenser, the capacities being obtained from the curve in fig. 2 for the values of 1000 × 8/8p obtained at

the various temperatures. In all these curves there is to be noticed a decided fall in capacity as the temperature rose; and the curves are practically straight lines.

The change of capacity for 1° rise in temperature, expressed as a percentage of the capacity at 0° C., obtained from the different curves is:

[merged small][merged small][ocr errors][merged small][merged small][ocr errors][merged small][ocr errors][merged small][ocr errors][merged small][merged small]

The dielectric of the condenser consisted partly of air and partly of water, the wires being out of the water for a considerable part of their length; and as the variation affected only the water, the total percentage change of capacity is less than that of the dielectric constant of the

water.

In the first two trials recorded above the water was placed in a large glass vessel 25 centimetres square, and the depth of water in it was about 15 centimetres; but in the other two the water was placed in a beaker 15 centimetres in diameter and the depth of water was about 15 centimetres : so in this case the air would have a greater relative share in the whole effect, and the observed change is less than in the former trials, where the variation in capacity is probably very nearly equal to the variation in the dielectric constant of water. In all four trials the wires were immersed to a depth of 5 or 6 centimetres, and were from 9 to 11 centimetres apart.

The assumption has been made that when the same. deflexion is obtained with the water- and air-condensers, the capacities are the same. By removing water from the vessel, and so decreasing the capacity, we can come into a position of resonance; and it was found that the maximum deflexions obtained when this was done were practically the same as the greatest deflexions obtained with the air-condenser, and so the conductivity of the water has no appreciable effect and the above assumption is probably permissible.

The specific conductivity of the water was measured, and found to be 3.7 × 10-6 at 18°.8 C.

These results show no indication of the large effect observed by Thwing at 4° C. With the exception of the results shown in fig. 5, all the observations are definitely lower at 4° than at 0°. According to Thwing the dielectric constant at 0° is 79.4 and at 4° 85.20, i. e. it increases by 7.2 per cent. In fig. 3 the capacity at 0° is 16:39; if it were 7.2 per cent. higher at 4°, the capacity would be 17.51 at that temperature.

An inspection of the curves shows that any change of that order could not fail to be detected.

In conclusion, I desire to express my best thanks to Professor Pollock for many valuable suggestions and for his continued encouragement during the work.

The University of Sydney.

LXXIII. On the Vapour-Pressure of Mercury at Ordinary Temperatures. By EDWARD W. MORLEY.

[ocr errors]

N 1890, when attempting to determine the density of hydrogen with accuracy, it became convenient to know the vapour-pressure of mercury at ordinary temperatures. There were extant two series of actual measurements at such temperatures, and four computations of the desired values from extrapolation formulæ founded on observations at higher temperatures.

The first of these computations was due to Regnault †, and was published by him in 1862. A second and a third were published by Hagen ‡ and by Hertz § in 1882. Ramsay and Young published the fourth in 1886.

Regnault also made a few observations at temperatures below 100°, which were published with those mentioned above. They seem to have been of service only in guiding conjecture as to the vapour-pressure which was assumed for 0°. Lastly, van der Plaats¶ in 1886 published direct determinations of the vapour-pressure at 0° and at the temperature of his laboratory.

The following table gives the vapour-pressure of mercury at certain temperatures according to these authorities :—

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][subsumed][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][subsumed][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]

*Communicated by the Author.

+ Relation des Expériences, ii. p. 506.
Wiedemann's Annalen, xvii. p. 618.

$ Ibid. xvii. p. 197.

Chem. Soc. Journal, xlix. p. 37.

¶ Recueil des Travaux Chimiques des Pays-Bas, v. p. 149.

0.008

These results, at first sight, seem to leave the whole matter very doubtful. The figures for 0°, for instance, differ in the ratio of 1, 25, 75, and 100.

The case is altered when we examine critically the different series of experiments. Regnault's values may first be dismissed from consideration. His value for 0° is simply assumed as differing from the truth by a negligible quantity; but the precision needed now is greater. Values for temperatures other than 0° depend on an interpolation formula. computed from values for 0°, 128°, 256°, 384°, and 512°. But Regnault had much difficulty with the observations at high temperatures, and the determinations were few and not concordant. Even if the true form of the function which expresses the relation between temperature and vapourpressure were known, Regnault's determinations would not, in the case of mercury, give values for the constants of the formula accurately enough for present needs; and he used only an empirical interpolation formula. Admirable as was his work, we have here to do with quantities which are smaller than the limits of accuracy which he claimed for such

measurements.

Hagen's measurements may also be disregarded. He determined the difference of level between the two arms of an exhausted syphon-gauge, one of which was connected with a vessel kept at a temperature at which the vapour-pressure of mercury may be considered negligible. He made numerous experiments as nearly as convenient to the temperatures of 0°, 50°, 100°, 150°, and 200°. By least square computations, he obtained the pressures corresponding to these precise temperatures; from which normal values an interpolation formula was calculated. The determinations for the two higher temperatures were known to be in error on account of the rapidity of the evaporation from the surface of the mercury. The work satisfactorily proved that the values given by Regnault below 100° were much too large. But Hagen's values cannot be accepted, if for no other reason, at least because his interpolation formula is too inconsistent with what we know of the behaviour of saturated vapours. The percentage increase of pressure due to an increase of temperature by ten degrees diminishes with increasing temperature; but Hagen's table gives an increase of 46 per cent. between 90° and 100°, while that from 0° to 10° is only 17 per cent.

The experiments of Hertz were made with care, and their principle was satisfactory. When a liquid evaporates into a gas whose pressure is greater than that of the saturated vapour of the liquid, the vapour very near the surface of the liquid

is very nearly saturated. If, at a given temperature, the liquid is brought to the same level in the two arms of a differential manometer containing only the saturated vapour of the liquid in one arm, and containing some gas in the other, a measurement of the pressure of the gas determines that of the saturated vapour. The uncertainty of the observations was 0.02 mm. An interpolation formula was computed from the observations at temperatures from 89° to 206°, and it would be difficult to improve his results until the very small quantities which represent the vapour-pressure of mercury at ordinary temperatures are made to depend, not on a somewhat remote extrapolation, and an extrapolation from values whose errors are many times as large as the quantities sought, but on direct measurement.

The measurements of Ramsay and Young are also very satisfactory for temperatures above 100°. The ratio of the absolute temperatures of water and of mercury having the same vapour-pressures varies so regularly, that if it is determined for a few temperatures, it is known for all temperatures. From this ratio and from the well-known vapour-pressures of water, they computed the vapour-pressure of mercury for temperatures from 135° to 520°. Below 135° the method could not be used, because the corresponding vapour-pressures of water are not well known; they therefore determined an extrapolation formula from the values for 160°, 220°, and 280°, from which they computed values for temperatures down to 40°.

We have, therefore, a determination by Hertz covering the interval from 0° to 100°, and one by Ramsay and Young for the interval from 40° to 100°. From 50° to 100° the mean difference between the two values is only 6.5 per cent. The mean difference between the values of the same experimenters from 120° to 220° is 8.6 per cent.; so that the agreement from 50° to 100° is satisfactory. But at 40° the difference is 27 per cent.

The observations of van der Plaats were made at the temperatures in question, and were numerous and careful. He passed a known volume of an inert gas through pure mercury in such a way as to saturate it with the vapour of mercury. The mercury was then collected by absorption and weighed. It is not easy to suggest a cause tending to make values obtained in this way larger than the truth. But it was and is impossible to accept them. The results at 0° and 10° are twenty-five and sixteen times as large as those of Hertz. Hertz found it impossible to detect the vapourpressure of mercury below 50°. But if the values of van der

« PreviousContinue »