260 
NATURE 
addition several auxiliary circulations have been constructed | 
in order to add new ranges of temperature to those already | 
obtainable by the main cycles, or at least to make certain | 
temperatures more practicable. Of these the principal 
are :—(1) a circulation of strong calcium chloride solution 
cooled to about —23° C. by methyl chloride, and then heated 
again to the required constant temperature in a thermostat ; 
(2) a nitrous oxide circulation, also in connection with the | 
methyl chloride, for temperatures about —go° C.; (3) air; 
(4) nitrogen; and (5) methane in connection with the oxygen 
cycle. In this way the entire range between +5° and 
—210° can be covered with but few gaps. 
The efficiency of the circulations is now very high, as 
each step in the cascade only lasts about 20 minutes, so that | 
liquid nitrogen even, which requires four operations, is 
obtained in 1 hour 20 minutes. For this purpose about 
6 kg. methyl chloride, 1-3 kg. ethylene, 2-5 kg. oxygen, and 
1-0 kg. nitrogen are required, with the pumps running at 
about 5, 6, 30, and 6 atmospheres pressure in the respective 
cycles. In order to obtain this speed of work a new large 
Burckhardt-Weiss pump, with a capacity of 1oo litres per 
second and capable of evacuating to 2 mm. of mercury, has 
been introduced before the vacuum cylinder of the Pictet 
conjugated pump in both methyl chloride and ethylene 
circulations. In the case when a bath of liquid nitrogen is 
required, one of the two Brotherhood compressors is put into 
the oxygen circulation, while the nitrogen is compressed in 
the modified Cailletet pump mentioned before, after passing 
through an auxiliary pump, to bring it to about 10 atmo- 
spheres pressure. 
Two methods are employed for conveying the liquid gas 
required from the cryogenic room to one of three experi- 
menting rooms. Where many measurements at a reduced 
temperature are desired, hence necessitating a frequent 
addition of liquid, methyl chloride, nitrous oxide, and 
ethylene have been conveyed in a copper tube well wrapped 
up in wool and paper for a distance of about 10 metres. 
In other cases the liquid is run into specially mounted 
vacuum glasses, which are transported to the experimental 
vessel, into which the liquid is siphoned over. Measure- 
ments have been made in this way with all the gases 
mentioned, in vessels which require the use of litres of 
liquid. 
One of Prof. Onnes’s chief objects in founding his cryo- 
genic department was the experimental determination of 
accurate isothermals of pure elementary gases and _ their 
binary mixtures, at temperatures near the critical. Although 
measurements had been made on methyl chloride, carbon 
dioxide, and mixtures of these with oxygen or hydrogen, 
no measurements on the latter gases alone had been 
attempted until about four years ago. The first necessity 
was a standard open mercury manometer for the very 
accurate determination of pressures up to 64 atmospheres. 
Since this has been available, measurements have been made 
on hydrogen, oxygen, and nitrogen at ordinary and at low 
temperatures to an accuracy unattained before in such work. 
At the same time very careful comparisons have been made 
between the expansion coefficients, at constant volume, of 
the same gases at low temperatures. In order that these 
measurements should be possible, it was also necessary to 
develop the system for the determination of temperature 
to the accuracy required. 
For this purpose use has been | 
made of constant volume hydrogen thermometers and of | 
specially prepared and calibrated platinum resistances or | 
thermoelements. 
For the most recent measurements it has been found 
advisable to employ, at the same time, both a temperature 
indicator and a temperature measurer. The - measuring 
instrument must be sensitive enough for the accuracy re- 
quired, but the indicator must be set to a higher order. A 
combination of a platinum resistance, wound upon glass 
and surrounding the experimental tube, with a thermo- 
element has usually been employed. 
The temperature of an evaporating liquid varies both 
from the want of homogeneity of the substance and of con- 
stancy in the pressure. By suitable adjustment of the 
pressure, it is possible to make these variations compensate 
one another, with the result that a nearly constant tempera- 
ture is obtained. In the thermoelectric circuit variations of 
o-o1 degree can be immediately seen and corrected by a suit- 
NO. 1785, VOL. 69] 
[JANUARY 14, 1904 
able small change of pressure. Variations due to the pump- 
ing machinery and other similar causes are corrected at 
once from the indications of an oil manometer. In addition, 
continual stirring by a small motor allows temperatures — 
down to —210° C. to be obtained, which remain constant 
and uniform to o.o1 degree for an hour in baths 15 cm. 
in depth. 
Preparations are now being made, by the addition of new 
pumps and other apparatus, for the installation, on the 
same principles, of a hydrogen cryostate, which will allow 
of the extension of the range of constant temperatures to 
the boiling point of hydrogen under reduced pressure. 
Since only minor technical difficulties are to be expected, 
it is probable that this cycle will be completed in a few 
years’ time. 
During the past decade several other important investi- 
gations have been made. Of these the best known is 
certainly the discovery and measurement by Dr. Zeeman 
of the effect which goes by his name. This was largely 
due to a set of careful experiments and measurements on 
Fic. 1.—The immediate apparatus for the preparation of liquid nitrogen 
or methane. The liquefied gas is being collected in the metal cylinders 
in middle of picture. This apparatus stands at the end of the room 
from which the view shown in 1896 was taken. 
kindred optical questions, and was tried several times with- 
out success. Immediately it was noticed Prof. Lorentz was 
able to supply the elementary theory, which has subsequently 
only been elaborated. The discovery thus belongs peculiarly 
to Leyden, while the partnership was fittingly acknowledged 
by the joint award of the Nobel prize in 1902. 
What Prof. Onnes considers as to some extent a new 
method of thermodynamical research has been recently de- 
veloped by him. It consists in making large accurate 
models of both the Gibbs (e, 7, v) and van der Waals 
(~, v, x) surfaces. By means of these many problems can 
be treated with comparative ease, which otherwise would 
either be impossible or would involve very lengthy calcula- 
tions. 
The following list of the principal researches from the 
beginning of the communications will show clearly the 
range of work covered. 
