274 
NALTORE 
[JANUARY 21, 1897 
liquid that rises higher than the other stands steadily 
at a convenient marked point, when the air-pump is kept 
vigorously going, with the lower air-pump stopcock closed. 
This marked point may be perhaps a few centimetres 
below the india-rubber collar, so as to allow the liquid 
surface ofit to be conveniently seen through a wide glass 
cylinder containing hot or cold water around it, applied 
to fulfil the thermal conditions referred to in Operation 
No. 6. In these present circumstances the vapour pres- 
sure is practically equal throughout the upper bent tube, 
and the portions of the glass tubes between its ends and 
the liquid surfaces in the two glass tubes. Hence the 
more volatile of the two liquids is kept cool at its surface 
by rapid evaporation, and the less volatile liquid is kept 
warm by rapid condensation of vapour into it, so that, 
by flow of vapour through the bent tube, the difference 
of temperatures required to equalise the vapour pressures 
is very nearly maintained. 
Operation No 5.—Close the upper three stopcocks, both 
air-pump stopcocks being already closed, and the two 
lowest metal stopcocks open. Leave the apparatus to 
itself until the temperatures become equalised. The 
difference of levels of the liquids in the two glass tubes, 
with proper corrections for their densities and for the 
difference, if any, of levels of the liquid surfaces in the 
two bottles, measures accurately the difference of vapour 
pressure over them, at the temperature to which they 
become equalised. 
Operation No.6.—Open the upper air-pump stopcock, 
work the air-pump and open the stopcock over the top of 
one of the two liquids for a minute or two and close it 
again. Do the same for the other liquid. Allow tem- 
peratures to be equalised to what they were at the end of 
Op. 5. If any air or other foreign volatile substance! 
has escaped from either liquid along with its proper 
vapour, its level will be seen higher than it was at the 
end of Op.5. The present operation (No. 6) must be 
continued long enough to distil out of either, or both 
liquids, any such foreign ingredients if, when originally 
introduced, any such impurity was contained. 
Operation No. 7.—By proper thermal appliances, in- 
dicated by the dotted lines in the diagram, and the lamp 
under the upper bent metal tube (inserted merely as an 
indication that somehow the metal tube is to be always 
slightly warmer than the warmer of the two liquid sur- 
faces, in order that there may be no condensation of 
vapour in it), bring the upper surfaces of the liquids to 
any other temperature, or to two different temperatures. 
The difference of levels of the liquids in the two tubes, 
with proper correction for the densities of the two liquids 
at their actual temperatures in different parts of their 
columns, gives the difference of vapour pressures for the 
actual temperatures of the two liquids at their upper 
surfaces. 
Operation No. 8.—TYo facilitate and approximately 
determine the hydrostatic correction for specific gravities 
at the actual temperatures of the two liquids, open wide 
the stopcocks above the tops of the two glass tubes, and 
let a little air run back from the air-pump, by very 
cautiously and slightly opening our upper air-pump stop- 
cock, and closing it again before the lower of the two 
liquid surfaces reaches the lower end of its glass tube. 
After that, by cautiously opening and closing our lower 
air-pump stopcock, let in a little air to the bottles until 
the mean level of the liquids in the two columns rises to 
nearly the same level as it had in the measured positions 
of Op. 5 or Op. 6. In the present circumstances, air in 
the upper bent metal tube resists diffusion of vapour 
through it sufficiently to prevent any important difference 
of temperatures from being produced by evaporation and 
condensation at the two liquid surfaces, and there is 
1 See Ostwald, ‘‘ Physico-Chemical Measurements,’ translated by Walker 
(Macmllian, 1892), last paragraph, page 112. 
NO! 1421, Vor 55 
i] 
practically perfect hydrostatic equilibrium of equal liquid 
pressures at the tops of the two columns. 
The vapour pressure of water is accurately known 
through a very wide range of temperature from Reg- 
nault’s experiments ; hence, if pure water be taken for 
one of our two liquids, the mode of experiment described 
above determines the vapour pressure of the other 
liquid. 
The apparatus may be kept day after day with the same 
liquids in it (all the stopcocks to be closed, except when 
it is not in use for observations) ; and thus, the observa- 
tions for difference of vapour pressures may he repeated 
day after day ; or a long series of observations may very 
easily be made to determine vapour pressures at different 
temperatures. Always before commencing observations, 
Operation 6 must be repeated to remove air or other im- 
purity, if any air has leaked in, or if air or other foreign 
volatile impurity has escaped from dissolution in either 
liquid into the vapour space above it. KELVIN. 
RELATIVE TEMPERATURES IN GEISSLER 
TUBES. 
| 
the Physical Institute of the Berlin University, Mr. 
R. W. Wood has been making a series of experi- 
ments, most interesting to students of astrophysics, with 
the object of investigating the relative temperatures at 
different parts of the discharge in a Geissler tube, with 
special reference to the stratification phenomena. Wiede- 
mann and Hittorf, and also the theoretical calculations 
of Warburg, have shown that the temperature of the gas 
in the positive part of the discharge lies far below red 
heat, while that of the negative light, according to 
Hittorf, is at least below the melting-point of platinum. 
These observations are for the most part corroborated 
by the experiments of Mr. Wood, who has investigated 
in this case a fixed part of the discharge in an atmo- 
sphere of nitrogen under varying pressures and currents 
of different strengths. The results obtained by employ- 
ing hydrogen instead of nitrogen established the fact 
that, under similar-conditions of pressure and strength 
of current, the heating was only about 11 per cent. of 
that found in the former case. It was found difficult, 
however, to keep a steady current with this gas. 
Perhaps more interesting are the results which he has 
been able to procure by determining the relative tem- 
peratures of the different parts of the space between the 
anode and kathode. For this he has designed a neat 
and very simple means, by which the positions of the 
bolometer inside the vacuum tube might be varied at 
will without impairing in the least degree the vacuum. 
The description of this apparatus will be found in 
the article in which the results of his observations 
have been published (Physical Review, November- 
December 1896, xxi.). We may, however, mention that 
the bolometer wire—that is, the wire which was placed 
in the different positions between the two poles of the 
Geissler tube to indicate the varying temperatures of the 
different parts of the discharge—was here composed of 
platino-iridium, and bent in the form of a loop. Its 
exact position could at any moment be read off from a 
vertical scale. It was thus found possible to make a 
complete map of the temperature changes inside the 
vacuum tube. 
In the unstratified anode light the temperature was 
sometimes constant for the greater part of the column, 
rising to a maximum near the middle, and falling off 
as the dark space was approached. The maximum was 
always found when the light was on the point of strati- 
fying, and sometimes at higher pressures. The exact 
conditions, however, could not be determined ; but the 
extent of the anode light played an important part. 
