Chapter II: Front Matter (2)

_Hydrogen._--Liquid hydrogen is the lightest liquid known to the
chemist, having a density slightly less than 0.07 as compared with
water, and being six times lighter than liquid marsh-gas, which is
next in order of lightness. One litre weighs only 70 grammes, and 1
gramme occupies a volume of 14-15 cc. In spite of its extreme
lightness, however, it is easily seen, has a well-defined meniscus and
drops well. At its boiling-point the liquid is only 55 times denser
than the vapour it is giving off, whereas liquid oxygen in similar
condition is 258 times denser than its vapour, and nitrogen 177 times.
Its atomic volume is about 14.3, that of liquid oxygen being 13.7,
and that of liquid nitrogen 16.6, at their respective boiling-points.
Its latent heat of vaporization about the boiling-point is about 121
gramme-calories, and the latent heat of fluidity cannot exceed 16
units, but may be less. Hydrogen appears to have the same specific
heat in the liquid as in the gaseous state, about 3.4. Its surface
tension is exceedingly low, about one-fifth that of liquid air at its
boiling-point, or one-thirty-fifth that of water at ordinary
temperatures, and this is the reason that bubbles formed in the liquid
are so small as to give it an opalescent appearance during ebullition.
The liquid is without colour, and gives no absorption spectrum.
Electric sparks taken in the liquid between platinum poles give a
spectrum showing the hydrogen lines C and F bright on a background of
continuous spectrum. Its refractive index at the boiling-point has
theoretically the value 1.11. It was measured by determining the
relative difference of focus for a parallel beam of light sent through
a spherical vacuum vessel filled successively with water, liquid
oxygen and liquid hydrogen; the result obtained was 1.12. Liquid
hydrogen is a non-conductor of electricity. The precise determination
of its boiling-point is a matter of some difficulty. The first results
obtained from the use of a platinum resistance thermometer gave -238°
C., while a similar thermometer made with an alloy of rhodium-platinum
indicated a value 8 degrees lower. Later, a gold thermometer indicated
about -249° C., while with an iron one the result was only -210° C. It
was thus evident that electrical resistance thermometers are not to be
trusted at these low temperatures, since the laws correlating
resistance and temperature are not known for temperatures at and below
the boiling-point of hydrogen, though they are certainly not the same
as those which hold good higher up the thermometric scale. The same
remarks apply to the use of thermo-electric junctions at such
exceptional temperatures. Recourse was therefore had to a
constant-volume hydrogen thermometer, working under reduced pressure,
experiments having shown that such a thermometer, filled with either a
simple or a compound gas (e.g. oxygen or carbonic acid) at an initial
pressure somewhat less than one atmosphere, may be relied upon to
determine temperatures down to the respective boiling-points of the
gases with which they are filled. The result obtained was -252° C.
Subsequently various other determinations were carried out in
thermometers filled with hydrogen derived from different sources, and
also with helium, the average value given by the experiments being
-252.5° C. (See "The Boiling Point of Liquid Hydrogen determined by
Hydrogen and Helium Gas Thermometers," _Proc. Roy. Soc._, 7th February
1901.) The critical temperature is about 30° absolute (-243° C.), and
the critical pressure about 15 atmospheres. Hydrogen has not only the
lowest critical temperature of all the old permanent gases, but it has
the lowest critical pressure. Given a sufficiently low temperature,
therefore, it is the easiest gas to liquefy so far as pressure is
concerned. Solid hydrogen has a temperature about 4° less. By
exhaustion under reduced pressure a still lower depth of cold may be
attained, and a steady temperature reached less than 16° above the
zero of absolute temperature. By the use of high exhaustion, and the
most stringent precautions to prevent the influx of heat, a
temperature of 13° absolute (-260° C.) may be reached. This is the
lowest steady temperature which can be maintained by the evaporation
of solid hydrogen. At this temperature the solid has a density of
about 0.077. Solid hydrogen presents no metallic characteristics, such
as were predicted for it by Faraday, Dumas, Graham and other chemists
and neither it nor the liquid is magnetic.

_The Approach to the Absolute Zero._--The achievement of Kamerlingh Onnes has brought about the realization of a temperature removed only 3° from the absolute zero, and the question naturally suggests itself whether there is any probability of a still closer approach to that point. The answer is that if, as is not impossible, there exists a gas, as yet unisolated, which has an atomic weight one-half that of helium, that gas, liquefied in turn by the aid of liquid helium, would render that approach possible, though the experimental difficulties of the operation would be enormous and perhaps prohibitive. The results of experiments bearing on this question and of theory based on them are shown in table II. The third column shows the critical temperature of the gas which can be liquefied by continuous expansion through a regenerative cooling apparatus, the operation being started from the initial temperature shown in the second column, while the fourth column gives the temperature of the resulting liquid. It will be seen that by the use of liquid or solid hydrogen as a cooling agent, it should be possible to liquefy a body having a critical temperature of about 6° to 8° on the absolute scale, and a boiling point of about 4° or 5°, while with the aid of liquid helium at an initial temperature of 5° we could liquefy a body having a critical temperature of 2° and a boiling point of 1°.

TABLE II.

+------------------+-------------+-------------+---------------+
| | Initial | Critical |Boiling Points.|
| Substance. | Temperature.|Temperature. | Abs. Degrees. |
| |Abs. Degrees.|Abs. Degrees.| |
+------------------+-------------+-------------+---------------+
| (Low red heat) | 760 | 304 | 195 (CO2) |
| (52° C.) | 325 | 130 | 86 (Air) |
| Liquid air under | | | |
| exhaustion | 75 | 30 | 20 (H) |
| Liquid hydrogen | 20 | 8 | 5 (He) |
| Solid hydrogen | 15 | 6 | 4 |
| Liquid helium | 5 | 2 | 1 |
+------------------+-------------+-------------+---------------+

It is to be remarked, however, that even so the physicist would not have attained the absolute zero, and he can scarcely hope ever to do so. It is true he would only be a very short distance from it, but it must be remembered that in a thermodynamic sense one degree low down the scale, say at 10° absolute, is equivalent to 30° at the ordinary temperature, and as the experimenter gets to lower and lower temperatures, the difficulties of further advance increase, not in arithmetical but in geometrical progression. Thus the step between the liquefaction of air and that of hydrogen is, thermodynamically and practically, greater than that between the liquefaction of chlorine and that of air, but the number of degrees of temperature that separates the boiling-points of the first pair of substances is less than half what it is in the case of the second pair. But the ratio of the absolute boiling-points in the first pair of substances is as 1 to 4, whereas in the second pair it is only 1 to 3, and it is this value that expresses the difficulty of the transition.

But though Ultima Thule may continue to mock the physicist's efforts, he will long find ample scope for his energies in the investigation of the properties of matter at the temperatures placed at his command by liquid air and liquid and solid hydrogen. Indeed, great as is the sentimental interest attached to the liquefaction of these refractory gases, the importance of the achievement lies rather in the fact that it opens out new fields of research and enormously widens the horizon of physical science, enabling the natural philosopher to study the properties and behaviour of matter under entirely novel conditions. We propose to indicate briefly the general directions in which such inquiries have so far been carried on, but before doing so will call attention to the power of absorbing gases possessed by cooled charcoal, which has on that account proved itself a most valuable agent in low temperature research.

TABLE III.--_Gas Absorption by Charcoal._

+---------------------------+-------------+-------------+
| | Volume | Volume |
| | absorbed at | absorbed at |
| | 0° Cent. | -185° Cent. |
+---------------------------+-------------+-------------+
| Helium | 2 cc. | 15 cc. |
| Hydrogen | 4 | 135 |
| Electrolytic gas | 12 | 150 |
| Argon | 12 | 175 |
| Nitrogen | 15 | 155 |
| Oxygen | 18 | 230 |
| Carbonic oxide | 21 | 190 |
| Carbonic oxide and oxygen | 30 | 195 |
+---------------------------+-------------+-------------+

_Gas Absorption by Charcoal._--Felix Fontana was apparently the first to discover that hot charcoal has the power of absorbing gases, and his observations were confirmed about 1770 by Joseph Priestley, to whom he had communicated them. A generation later Theodore de Saussure made a number of experiments on the subject, and noted that at ordinary temperatures the absorption is accompanied with considerable evolution of heat. Among subsequent investigators were Thomas Graham and Stenhouse, Faure and Silberman, and Hunter, the last-named showing that charcoal made from coco-nut exhibits greater absorptive powers than other varieties. In 1874 Tait and Dewar for the first time employed charcoal for the production of high vacua, by using it, heated to a red heat, to absorb the mercury vapour in a tube exhausted by a mercury pump; and thirty years afterwards it occurred to the latter investigator to try how its absorbing powers are affected by cooling it, with the result that he found them to be greatly enhanced. Some of his earlier observations are given in table III., but it must be pointed out that much larger absorptions were obtained subsequently when it was found that the quality of the charcoal was greatly influenced by the mode in which it was prepared, the absorptive power being increased by carbonizing the coco-nut shell slowly at a gradually increasing temperature. The results in the table were all obtained with the same specimen of charcoal, and the volumes of the gases absorbed, both at ordinary and at low temperatures, were measured under standard conditions--at 0° C., and 760 mm. pressure. It appears that at the lower temperature there is a remarkable increase of absorption for every gas, but that the increase is in general smaller as the boiling-points of the various gases are lower. Helium is conspicuous for the fact that it is absorbed to a comparatively slight extent at both the higher and the lower temperature, but in this connexion it must be remembered that, being the most volatile gas known, it is being treated at a temperature which is relatively much higher than the other gases. At -185° (= 88° abs.), while hydrogen is at about 4½ times its boiling-point (20° abs.), helium is at about 20 times its boiling-point (4.5° abs.), and it might, therefore, be expected that if it were taken at a temperature corresponding to that of the hydrogen, i.e. at 4 or 5 times its boiling-point, or say 20° abs., it would undergo much greater absorption. This expectation is borne out by the results shown in table IV., and it may be inferred that charcoal cooled in liquid helium would absorb helium as freely as charcoal cooled in liquid hydrogen absorbs hydrogen. It is found that a given specimen of charcoal cooled in liquid oxygen, nitrogen and hydrogen absorbs about equal volumes of those three gases (about 260 cc. per gramme); and, as the relation between volume and temperature is nearly lineal at the lowest portions of either the hydrogen or the helium absorption, it is a legitimate inference that at a temperature of 5° to 6° abs. helium would be as freely absorbed by charcoal as hydrogen is at its boiling-point and that the boiling-point of helium lies at about 5° abs.

TABLE IV.--_Gas Absorption by Charcoal at Low Temperatures._

+--------------------------------------------+--------+----------+
| | Helium.| Hydrogen.|
| Temperature. |Vols. of| Vols. of |
| |Carbon. | Carbon. |
+--------------------------------------------+--------+----------+
| -185° C. (boiling-point of liquid air) | 2½ | 137 |
| -210° C. (liquid air under exhaustion) | 5 | 180 |
| -252° C. (boiling-point of liquid hydrogen)| 160 | 258 |
| -258° C. (solid hydrogen) | 195 | .. |
+--------------------------------------------+--------+----------+

The rapidity with which air is absorbed by charcoal at -185° C. and under small pressures is illustrated by table V., which shows the reductions of pressure effected in a tube of 2000 cc. capacity by means of 20 grammes of charcoal cooled in liquid air.

TABLE V.--_Velocity of Absorption._

+-----------+--------+-----------+---------+
| Time of |Pressure| Time of |Pressure |
|Exhaustion.| in mm. |Exhaustion.| in mm. |
+-----------+--------+-----------+---------+
| 0 sec. | 2.190 | 60 sec. | 0.347 |
| 10 " | 1.271 | 2 min. | 0.153 |
| 20 " | 0.869 | 5 " | 0.0274 |
| 30 " | 0.632 | 10 " | 0.00205 |
| 40 " | 0.543 | 19 " | 0.00025 |
| 50 " | 0.435 | .. | .. |
+-----------+--------+-----------+---------+

TABLE VI.

+----------+----------+-----------+
| Volume | Occlusion| Occlusion |
| of Gas | Hydrogen | Nitrogen |
| absorbed.| Pressure.| Pressure. |
+----------+----------+-----------+
| cc. | mm. | mm. |
| 0 | 0.00003 | 0.00005 |
| 5 | 0.0228 | .. |
| 10 | 0.0455 | .. |
| 15 | 0.0645 | .. |
| 20 | 0.0861 | .. |
| 25 | 0.1105 | .. |
| 30 | 0.1339 | 0.00031 |
| 35 | 0.1623 | .. |
| 40 | 0.1870 | .. |
| 130 | .. | 0.00110 |
| 500 | .. | 0.00314 |
| 1000 | .. | 0.01756 |
| 1500 | .. | 0.02920 |
| 2500 | .. | 0.06172 |
+----------+----------+-----------+

_Charcoal Occlusion Pressures._--For measuring the gas concentration,
pressure and temperature, use may be made of an apparatus of the type
shown in fig. 5. A mass of charcoal, E, immersed in liquid air, is
employed for the preliminary exhaustion of the McLeod gauge G and of
the charcoal C, which is to be used in the actual experiments, and is
then sealed off at S. The bulb C is then placed in a large spherical
vacuum vessel containing liquid oxygen which can be made to boil at
any definite temperature under diminished pressure which is measured
by the manometer R. The volume of gas admitted into the charcoal is
determined by the burette D and the pipette P, and the corresponding
occlusion pressure at any concentration and any temperature below 90°
abs. by the gauge G. In presence of charcoal, and for small
concentrations, great variations are shown in the relation between the
pressure and the concentration of different gases, all at the same
temperature. Table VI. gives the comparison between hydrogen and
nitrogen at the temperature of liquid air, 25 grammes of charcoal
being employed. It is seen that 15 cc. of hydrogen produce nearly the
same pressure (0.0645 mm.) as 2500 cc. of nitrogen (0.06172 mm.). This
result shows how enormously greater, at the temperature of liquid air,
is the volatility of hydrogen as compared with that of nitrogen. In
the same way the concentrations, for the same pressure, vary greatly
with temperature, as is exemplified by table VII., even though the
pressures are not quite constant. The temperatures employed were the
boiling-points of hydrogen, oxygen and carbon dioxide.

TABLE VII.

+---------------+---------------+--------+-----------+
| | Concentration |Pressure|Temperature|
| Gas. |in cc. per grm.| in mm. | Absolute. |
| | of Charcoal. | | |
+---------------+---------------+--------+-----------+
| Helium | 97 | 2.2 | 20° |
| Hydrogen | 397 | 2.2 | 20° |
| Hydrogen | 15 | 2.1 | 90° |
| Nitrogen | 250 | 1.6 | 90° |
| Oxygen | 300 | 1.0 | 90° |
| Carbon dioxide| 90 | 3.6 | 195° |
+---------------+---------------+--------+-----------+

TABLE VIII.

+---------------+-------------+------------+------------+
| |Concentration| Molecular | Mean |
| Gas. | cc. per grm.|Latent Heat.|Temperature.|
| | | | Absolute. |
+---------------+-------------+------------+------------+
| Helium | 97 | 483.0 | 18° |
| Hydrogen | 390 | 524.4 | 18° |
| Hydrogen | 20 | 2005.6 | 78° |
| Nitrogen | 250 | 3059.0 | 82° |
| Oxygen | 300 | 3146.4 | 82° |
| Carbon dioxide| 90 | 6099.6 | 180° |
+---------------+-------------+------------+------------+

_Heat of Occlusion._--In every case when gases are condensed to the
liquid state there is evolution of heat, and during the absorption of
a gas in charcoal or any other occluding body, as hydrogen in
palladium, the amount of heat evolved exceeds that of direct
liquefaction. From the relation between occlusion-pressure and
temperature at the same concentration, the reaction being reversible,
it is possible to calculate this heat evolution. Table VIII. gives the
mean molecular latent heats of occlusion resulting from Dewar's
experiments for a number of gases, having concentrations in the
charcoal as shown. The concentrations were so regulated as to start
with an initial pressure not exceeding 3 mm. at the respective
boiling-points of hydrogen, nitrogen, oxygen and carbon dioxide.

_Production of High Vacua._--Exceedingly high vacua can be obtained by the aid of liquid gases, with or without charcoal. If a vessel containing liquid hydrogen be freely exposed to the atmosphere, a rain of snow (solid air) at once begins to fall upon the surface of the liquid; similarly, if one end of a sealed tube containing ordinary air be immersed in the liquid, the same thing happens, but since there is now no new supply to take the place of the air that has been solidified and has accumulated in the cooled portion of the tube, the pressure is quickly reduced to something like one-millionth of an atmosphere, and a vacuum is formed of such tenuity that the electric discharge can be made to pass only with difficulty. Liquid air can be employed in the same manner if the tube, before sealing, is filled with some less volatile gas or vapour, such as sulphurous acid, benzol or water vapour. But if a charcoal condenser be used in conjunction with the liquid air it becomes possible to obtain a high vacuum when the tube contains air initially. For instance, in one experiment, with a bulb having a capacity of 300 cc. and filled with air at a pressure of about 1.7 mm. and at a temperature of 15° C., when an attached condenser with 5 grammes of charcoal was cooled in liquid air, the pressure was reduced to 0.0545 mm. of mercury in five minutes, to 0.01032 mm. in ten minutes, to 0.000139 mm. in thirty minutes, and to 0.000047 mm. in sixty minutes. The condenser then being cooled in liquid hydrogen the pressure fell to 0.0000154 mm. in ten minutes, and to 0.0000058 mm. in a further ten minutes when solid hydrogen was employed as the cooling agent, and no doubt, had it not been for the presence of hydrogen and helium in the air, an even greater reduction could have been effected. Another illustration of the power of cooled charcoal to produce high vacua is afforded by a Crookes radiometer. If the instrument be filled with helium at atmospheric pressure and a charcoal bulb attached to it be cooled in liquid air, the vanes remain motionless even when exposed to the concentrated beam of an electric arc lamp; but if liquid hydrogen be substituted for the liquid air rapid rotation at once sets in. When a similar radiometer was filled with hydrogen and the attached charcoal bulb was cooled in liquid air rotation took place, because sufficient of the gas was absorbed to permit motion. But when the charcoal was cooled in liquid hydrogen instead of in liquid air, the absorption increased and consequently the rarefaction became so high that there was no motion when the light from the arc was directed on the vanes. These experiments again permit of an inference as to the boiling-point of helium. A fall of 75% in the temperature of the charcoal bulb, from the boiling-point of air to the boiling-point of hydrogen, reduced the vanes to rest in the case of the radiometer filled with hydrogen; hence it might be inferred that a fall of like amount from the boiling-point of hydrogen would reduce the vanes of the helium radiometer to rest, and consequently that the boiling-point of helium would be about 5° abs.

The vacua obtainable by means of cooled charcoal are so high that it is difficult to determine the pressures by the McLeod gauge, and the radiometer experiments referred to above suggested the possibility of another means of ascertaining such pressures, by determining the pressures below which the radiometer would not spin. The following experiment shows how the limit of pressure can be ascertained by reference to the pressures of mercury vapour which have been very accurately determined through a wide range of temperature. To a radiometer (fig. 6) with attached charcoal bulb B was sealed a tube ending in a small bulb A containing a globule of mercury. The radiometer and bulb B were heated, exhausted and repeatedly washed out with pure oxygen gas, and then the mercury was allowed to distil for some time into the charcoal cooled in liquid air. On exposure to the electric beam the vanes began to spin, but soon ceased when the bulb A was cooled in liquid air. When, however, the mercury was warmed by placing the bulb in liquid water, the vanes began to move again, and in the particular radiometer used this was found to happen when the temperature of the mercury had risen to -23° C. corresponding to a pressure of about one fifty-millionth of an atmosphere.

For washing out the radiometer with oxygen the arrangement shown in fig. 7 is convenient. Here A is a bulb containing perchlorate of potash, which when heated gives off pure oxygen; C is again the radiometer and B the charcoal bulb. The side tube E is for the purpose of examining the gas given off by minerals like thorianite or the gaseous products of the transformation of radioactive bodies.

_Analytic Uses._--Another important use of liquid gases is as analytic agents, and for this purpose liquid air is becoming an almost essential laboratory reagent. It is one of the most convenient agents for drying gases and for their purification. If a mixture of gases be subjected to the temperature of liquid air, it is obvious that all the constituents that are more condensable than air will be reduced to liquid, while those that are less condensable will either remain as a gaseous residue or be dissolved in the liquid obtained. The bodies present in the latter may be separated by fractional distillation, while the contents of the gaseous residue may be further differentiated by the air of still lower temperatures, such as are obtainable by liquid hydrogen. An apparatus such as the following can be used to separate both the less and the more volatile gases of the atmosphere, the former being obtained from their solution in liquid air by fractional distillation at low pressure and separation of the condensable part of the distillate by cooling in liquid hydrogen, while the latter are extracted from the residue of liquid air, after the distillation of the first fraction, by allowing it to evaporate gradually at a temperature rising only very slowly.

In fig. 8, A represents a vacuum-jacketed vessel, containing liquid
air; this can be made to boil at reduced pressure and therefore be
lowered in temperature by means of an air-pump, which is in
communication with the vessel through the pipe _s_. The liquid boiled
away is replenished when necessary from the reservoir C, _p_ being a
valve, worked by handle _q_, by which the flow along _r_ is regulated.
The vessel B, immersed in the liquid air of A, communicates with the
atmosphere by _a_; hence when the temperature of A falls under
exhaustion below that of liquid air, the contents of B condense, and
if the stop-cock _m_ is kept open, and _n_ shut, air from the outside
is continuously sucked in until B is full of liquid, which contains in
solution the whole of the most volatile gases of the atmosphere which
have passed in through _a_. At this stage of the operation _m_ is
closed and _n_ opened, a passage thus being opened along _b_ from A to
the remainder of the apparatus seen on the left side of the figure.
Here E is a vacuum vessel containing liquid hydrogen, and _d_ a
three-way cock by which communication can be established either
between _b_ and D, between _b_ and _e_, the tube leading to the
sparking-tube _g_, or between D and _e_. If now _d_ is arranged so
that there is a free passage from _b_ to D, and the stop-cock _n_ also
opened, the gas dissolved in the liquid in B, together with some of
the most volatile part of that liquid, quickly distils over into D,
which is at a much lower temperature than B, and some of it condenses
there in the solid state. When a small fraction of the contents of B
has thus distilled over, _d_ is turned so as to close the passage
between D and _b_ and open that between D and _e_, with the result
that the gas in D is pumped out by the mercury-pump, shown
diagrammatically at _F_, along the tube _e_ (which is immersed in the
liquid hydrogen in order that any more condensable gas carried along
by the current may be frozen out) to the sparking-tube or tubes _g_,
where it can be examined spectroscopically. When the apparatus is used
to separate the least volatile part of the gases in the atmosphere,
the vessel E and its contents are omitted, and the tube _b_ made to
communicate with the pump through a number of sparking-tubes which can
be sealed off successively. The nitrogen and oxygen which make up the
bulk of the liquid in B are allowed to evaporate gradually, the
temperature being kept low so as to check the evaporation of gases
less volatile than oxygen. When most of the oxygen and nitrogen have
thus been removed, the stop-cock _n_ is closed, and the tubes
partially exhausted by the pump; spectroscopic examination is made of
the gases they contain, and repeated from time to time as more gas is
allowed to evaporate from B. The general sequence of spectra, apart
from those of nitrogen, oxygen and carbon compounds, which are never
eliminated by the process of distillation alone, is as follows: The
spectrum of argon first appears, followed by the brightest (green and
yellow) rays of krypton. Then the intensity of the argon spectrum
wanes and it gives way to that of krypton, until, as Runge observed,
when a Leyden jar is in the circuit, the capillary part of the
sparking-tube has a magnificent blue colour, while the wide ends are
bright pale yellow. Without a jar the tube is nearly white in the
middle and yellow about the poles. As distillation proceeds, the
temperature of the vessel containing the residue of liquid air being
allowed to rise slowly, the brightest (green) rays of xenon begin to
appear, and the krypton rays soon die out, being superseded by those
of xenon. At this stage the capillary part of the sparking-tube is,
with a jar in circuit, a brilliant green, and it remains green, though
less brilliant, if the jar is removed.

An improved form of apparatus for the fractionation is represented in
fig. 9. The gases to be separated, that is, the least volatile part of
atmospheric air, enter the bulb B from a gasholder by the tube _a_
with stop-cock _c_. B, which is maintained at a low temperature by
being immersed in liquid hydrogen, A, boiling under reduced pressure,
in turn communicates through the tube _b_ and stop-cock _d_ with a
sparking-tube or tubes _f_, and so on through _e_ with a mercurial
pump. To use the apparatus, stop-cock _d_ is closed and _c_ opened,
and gas allowed to pass from the gasholder into B, where it is
condensed in the solid form. Stop-cock _c_ then being closed and _d_
opened, gas passes into the exhausted tube _f_, where it is examined
with the spectroscope. The vessel D contains liquid air, in which the
tube _e_ is immersed in order to condense vapour of mercury which
would otherwise pass from the pump into the sparking-tube. The success
of the operation of separating all the gases which occur in air and
which boil at different temperatures, depends on keeping the
temperature of B as low as possible, as will be understood from the
following consideration:--

The pressure _p_, of a gas G, above the same material in the liquid
state, at temperature T, is given approximately by the formula

B
log p = A - ---,
T

where A and B are constants for the same material. For some other gas
G´ the formula will be

B1
log p1 = A1 - ---,
T

and

p B1 - B
log --- = A - A1 + ------,
p1 T

Now for argon, krypton and xenon respectively the values of A are
6.782, 6.972 and 6.963, and those of B are 339, 496.3 and 669.2; so
that for these substances and many others A - A1 is always a small
quantity, while (B1 - B)/T is considerable and increases as T
diminishes. Hence the ratio of _p_ to _p_1 increases rapidly as T
diminishes, and by evaporating all the gases from the solid state, and
keeping the solid at as low a temperature as possible, the gas that is
taken off by the mercurial pump first consists mainly of the substance
which has the lowest boiling point, in this case nitrogen, and is
succeeded with comparative abruptness by the gas which has the next
higher boiling point. Examination of the spectrum in the sparking-tube
easily reveals the change from one gas to another, and when that is
observed the reservoirs into which the gases are pumped can be changed
and the fractions stored separately. Or several sparking-tubes may be
arranged so as to form parallel communications between _b_ and _e_,
and can be successively sealed off at the desired stages of
fractionation.

Analytical operations can often be performed still more conveniently with the help of charcoal, taking advantage of the selective character of its absorption, the general law of which is that the more volatile the gas the less is it absorbed at a given temperature. The following are some examples of its employment for this purpose. If it be required to separate the helium which is often found in the gases given off by a thermal spring, they are subjected to the action of charcoal cooled with liquid air. The result is the absorption of the less volatile constituents, i.e. all except hydrogen and helium. The gaseous residue, with the addition of oxygen, is then sparked, and the water thus formed is removed together with the excess of oxygen, when helium alone remains. Or the separation may be effected by a method of fractionation as described above. To separate the most volatile constituents of the atmosphere an apparatus such as that shown in fig. 10 may be employed. In one experiment with this, when 200 c.c. was supplied from the graduated gas-holder F to the vessel D, containing 15 grammes of charcoal cooled in liquid air, the residue which passed on unabsorbed to the sparking-tube AB, which had a small charcoal bulb C attached, showed the C and F lines of hydrogen, the yellow and some of the orange lines of neon and the yellow and green of helium. By using a second charcoal vessel E, with stop-cocks at H, I, J, K and L to facilitate manipulation, considerable quantities of the most volatile gases can be collected. After the charcoal in E has been saturated, the stop-cock K is closed and I and J are opened for a short time, to allow the less condensable gas in E to be sucked into the second condenser D along with some portion of air. The condenser E is then taken out of the liquid air, heated quickly to 15° C. to expel the occluded air and replaced. More air is then passed in, and by repeating the operation several times 50 litres of air can be treated in a short time, supplying sparking-tubes which will show the complete spectra of the volatile constituents of the air.

The less volatile constituents of the atmosphere, krypton and xenon, may be obtained by leading a current of air, purified by passage through a series of tubes cooled in liquid air, through a charcoal condenser also cooled in liquid air. The condenser is then removed and placed in solid carbon dioxide at -78° C. The gas that comes off is allowed to escape, but what remains in the charcoal is got out by heating and exhaustion, the carbon compounds and oxygen are removed and the residue, consisting of nitrogen with krypton and xenon, is separated into its constituents by condensation and fractionation. Another method is to cover a few hundred grammes of charcoal with old liquid air, which is allowed to evaporate slowly in a silvered vacuum vessel; the gases remaining in the charcoal are then treated in the manner described above.

Charcoal enables a mixture containing a high percentage of oxygen to be extracted from the atmosphere. In one experiment 50 grammes of it, after being heated and exhausted were allowed to absorb air at -185° C.; some 5 or 6 litres were taken up in ten minutes, and it then presumably contained air of the composition of the atmosphere, i.e. 20% oxygen and 80% nitrogen, as shown in fig. 11. But when more air was passed over it, the portion that was not absorbed was found to consist of about 98% nitrogen, showing that excess of oxygen was being absorbed, and in the course of a few hours the occluded gas attained a new and apparently definite composition exhibited in fig. 12. When the charcoal containing this mixture was transferred to a vacuum vessel and allowed to warm up slowly, the successive litres of gas when collected and analyzed separately showed the following composition:--

1st litre 18.5% oxygen
2nd litre 20.6% "
3rd litre 53.0% "
4th litre 72.0% "
5th litre 79.0% "
6th litre 84.0% "

TABLE IX.

+----------------+--------+--------------+-----------+-------------+
| | | Liquid Volume| |Volume of Gas|
| |Boiling | of 1 gram at |Latent Heat|at 0° C. and |
| Liquid Gases. | Point. | Boiling Point| in gram | 760 mm. per |
| | | in c.c. | Calories. | gram Calorie|
| | | | | in c.c. |
+----------------+--------+--------------+-----------+-------------+
| Sulphurous acid| + 10°C.| 0.7 | 97.0 | 3.6 |
| Carbonic acid | - 78.0 | 0.65 (solid)| 142.4 | 3.6 |
| Ethylene | -103.0 | 1.7 | 119.0 | 7.0 |
| Oxygen | -182.5 | 0.9 | 53.0 | 13.2 |
| Nitrogen | -195.6 | 1.3 | 50.0 | 15.9 |
| Hydrogen | -252.5 | 14.3 | 125.0 | 88.9 |
| Helium | -269.0 | 7.0 | 13.0 | 450.0 |
+----------------+--------+--------------+-----------+-------------+

_Calorimetry._--Certain liquid gases lend themselves conveniently to the construction of a calorimeter, in which the heat in weighed quantities of any substance with which it is desired to experiment may be measured by the quantity of liquid gas they are able to evaporate. One advantage of this method is that a great range of temperature is available when liquid air, oxygen, nitrogen or hydrogen is employed as the calorimetric substance. Another is the relatively large quantity of gas yielded by the evaporation, as may be seen from table IX., which shows the special physical constants of the various gases that are of importance in calorimetry. In consequence it is easy to detect 1/50 gram calorie with liquid air and so little as 1/300 gram calorie with liquid hydrogen.

The apparatus (fig. 13) consists of a large vacuum vessel A, of 2 or 3
litres' capacity, containing liquid air, in which is inserted a
smaller vacuum vessel B, of 25-30 c.c. capacity, having sealed to it a
long narrow tube G that projects above the mouth of A and is held in
place by some loosely packed cotton wool. To the top of this tube the
test tube C, containing the material under investigation, is connected
by a piece of flexible rubber tubing D; this enables C to be tilted so
as to throw a piece or pieces of the contained material into the
calorimeter. An improved form of this receptacle, attached to B by a
flexible tube at D´, is shown at C´. In this P is a wire movable
through a cork Q and having at its end a hook by which a piece of the
substance under examination can be pulled up and dropped into B. In
the absence of other arrangements the substance is at the temperature
of the room, but when lower initial temperatures are desired a vacuum
vessel H containing solid carbonic acid, liquid ethylene, air or other
gas, can be placed to envelop C or C´, or higher temperatures may be
obtained by filling the surrounding vessel with vapour of water or
other liquids. The gas volatilized in B is conveyed by a side tube E
to be collected in a graduated receiver F over water, oil or other
liquid. If liquid hydrogen is to be used as the calorimetric substance
the instrument must be so modified as to prevent the ordinary
atmosphere from entering G, and to that end a current of hydrogen
supplied from a Kipp apparatus is arranged to flow continuously
through D and E until the moment of making the experiment, when it is
cut off by a suitable stop-cock. In this case the outer vessel must
contain liquid hydrogen instead of liquid air.

TABLE X.

+-----------+---------+----------+----------+
| | 18° to | -78° to | -188° to |
| Substance.| -78° C.,| -188° C.,| -252° C.,|
| | or, at | or, at | or, at |
| | -30° C. | -133° C. | -220° C. |
+-----------+---------+----------+----------+
| Diamond | 0.0794 | 0.0190 | 0.0043 |
| Graphite | 0.1341 | 0.0599 | 0.0133 |
| Ice | 0.463* | 0.285 | 0.146 |
+-----------+---------+----------+----------+

* This is from -18° to -78° in the ice experiment.

Dewar used pure metallic lead for the purpose of conveying definite amounts of heat to liquid gas calorimeters of this kind, that metal being selected on the ground of the small variation in its specific heat at low temperatures. He was thus able to determine the latent heats of evaporation of liquid oxygen, nitrogen and hydrogen directly at their boiling points, and he also ascertained the specific heats of a large number of inorganic and organic bodies, and of some gases in the solid state, such as carbon dioxide, sulphurous acid and ammonia. Perhaps his most interesting results were those which showed the variation in the specific heats of diamond, graphite and ice as typical bodies (table X.). With Professor Curie he used both the liquid oxygen and the liquid hydrogen calorimeter for preliminary measurements of the rate at which radium bromide gives out energy at low temperatures. The quantity of the salt available was 0.42 gram, and the thermal evolutions were as follows:--

Gas evolved Calories
per minute. per hour.

Liquid oxygen 5.5 cc. 22.8 \
Liquid hydrogen 51.0 " 31.6 > Crystals.
Melting ice .. 24.1 /
Liquid oxygen 2.0 " 8.3 After fusion.
Liquid oxygen 2.5 " 10.3 Emanation condensed.

The apparent increase of heat evolution at the temperature of liquid hydrogen was probably due to the calorimeter being too small; hydrogen spray was thus carried away with the gas, making the volume of gas too great and inferentially also the heat evolved.

Liquid air and liquid hydrogen calorimeters open up an almost unlimited field of research in the determination of specific heats and other thermal constants, and are certain to become common laboratory instruments for such purposes.

_Chemical Action._--By extreme cold chemical action is enormously reduced, though it may not in all cases be entirely abolished even at the lowest temperatures yet attained; one reason for this diminution of activity may doubtless be sought in the fact that in such conditions most substances are solid, that is, in the state least favourable to chemical combination. Thus an electric pile of sodium and carbon ceases to yield a current when immersed in liquid oxygen. Sulphur, iron and other substances can be made to burn under the surface of liquid oxygen if the combustion is properly established before the sample is immersed, and the same is true of a fragment of diamond. Nitric oxide in the gaseous condition combines instantly with free oxygen, producing the highly-coloured gas, nitric peroxide, but in the solid condition it may be placed in contact with liquid oxygen without showing any signs of chemical action. If the combination of a portion of the mixture is started by elevation of temperature, then detonation may take place throughout the cooled mass. The stability of endothermic bodies like nitric oxide and ozone at low temperatures requires further investigation. The behaviour of fluorine, which may be regarded as the most active of the elements, is instructive in this respect. As a gas, cooled to -180° C. it loses the power of attacking glass; similarly silicon, borax, carbon, sulphur and phosphorus at the same temperature do not become incandescent in an atmosphere of the gas. Passed into liquid oxygen, the gas dissolves and imparts a yellowish tint to the liquid; if the oxygen has been exposed to the air for some hours, the fluorine produces a white flocculent precipitate, which if separated by filtering deflagrates with violence as the temperature rises. It appears to be a hydrate of fluorine. As a liquid at -210° fluorine attacks turpentine also cooled to that temperature with explosive force and the evolution of light, while the direction of a jet of hydrogen upon its surface is immediately followed by combination and a flash of flame. Even when the point of a tube containing solid fluorine is broken off under liquid hydrogen, a violent explosion ensues.

_Photographic Action._--The action of light on photographic plates, though greatly diminished at -180°, is far from being in abeyance; an Eastman film, for instance, remains fairly sensitive at -210°. At the still lower temperature of liquid hydrogen the photographic activity is reduced to about half what it is at that of liquid air; in other words, about 10% of the original sensitivity remains. Experiments carried out with an incandescent lamp, a Röntgen bulb and the ultra-violet spark from magnesium and cadmium, to discover at what distances from the source of light the plates must be placed in order to receive an equal photographic impression, yielded the results shown in table XI.

TABLE XI.

+--------------------+---------+----------+--------------+
| | Cooled | Uncooled | Ratio of |
| Source of Light. | Plate. | Plate. | Intensities |
| | | | at Balance. |
+--------------------+---------+----------+--------------+
| 16 C.P. lamp | 20 in. | 50 in. | 1 to 6 |
| Röntgen bulb | 10 in. | 24¾ in. | 1 to 6 |
| Ultra-violet spark | 22½ in. | 90 in. | 1 to 16 |
+--------------------+---------+----------+--------------+

It appears that the photographic action of both the incandescent lamp and the Röntgen rays is reduced by the temperature of liquid air to 17% of that exerted at ordinary temperatures, while ultra-violet radiation retains only 6%. It is possible that the greater dissipation of the latter by the photographic film at low temperatures than at ordinary ones is due to its absorption and subsequent emission as a phosphorescent glow, and that if the plate could be developed at a low temperature it would show no effect, the photographic action taking place subsequently through an internal phosphorescence in the film during the time it is heating up. With regard to the transparency of bodies to the Röntgen radiation at low temperatures, small tubes of the same bore, filled with liquid argon and chlorine, potassium, phosphorus, aluminium, silicon and sulphur, were exposed at the temperature of liquid air (in order to keep the argon and chlorine solid), in front of a photographic plate shielded with a sheet of aluminium, to an X-ray bulb. The sequence of the elements as mentioned represents the order of increasing opacity observed in the shadows. Sodium and liquid oxygen and air, nitrous and nitric oxides, proved much more transparent than chlorine. Tubes of potassium, argon and liquid chlorine showed no very marked difference of density on the photographic plates. It appears that argon is relatively more opaque to the Röntgen radiation than either oxygen, nitrogen or sodium, and is on a level with potassium, chlorine, phosphorus, aluminium and sulphur. This fact may be regarded as supporting the view that the atomic weight of argon is twice its density relative to hydrogen, since in general the opacity of elements in the solid state increases with the atomic weight.

_Phosphorescence._--Phosphorescing sulphides of calcium, which are luminous at ordinary temperatures, and whose emission of light is increased by heating, cease to be luminous if cooled to -80° C. But their light energy is merely rendered latent, not destroyed, by such cold, and they still retain the capacity of taking in light energy at the low temperature, to be evolved again when they are warmed. At the temperature of liquid air many bodies become phosphorescent which do not exhibit the phenomenon at all, or only to a very slight extent, at ordinary temperatures, e.g. ivory, indiarubber, egg-shells, feathers, cottonwool, paper, milk, gelatine, white of egg, &c. Of definite chemical compounds, the platinocyanides among the inorganic bodies seem to yield the most brilliant effects. Crystals of ammonium platinocyanide, if stimulated by exposure to the ultra-violet radiation of the electric arc--or better still of a mercury vapour lamp in quartz--while kept moistened with liquid air, may be seen in the dark to glow faintly so long as they are kept cold, but become exceedingly brilliant when the liquid air evaporates and the temperature rises. Among organic bodies the phenomenon is particularly well marked with the ketonic compounds and others of the same type. The chloro-, bromo-, iodo-, sulpho- and nitro-compounds show very little effect as a rule. The activity of the alcohols, which is usually considerable, is destroyed by the addition of a little iodine. Coloured salts, &c., are mostly inferior in activity to white ones. When the lower temperature of liquid hydrogen is employed there is a great increase in phosphorescence under light stimulation as compared with that observed with liquid air. The radio-active bodies, like radium, which exhibit self-luminosity in the dark, maintain that luminosity unimpaired when cooled in liquid hydrogen.

Some crystals become for a time self-luminous when placed in liquid hydrogen, because the high electric stimulation due to the cooling causes actual electric discharges between the crystal molecules. This phenomenon is very pronounced with nitrate of uranium and some platinocyanides, and cooling such crystals even to the temperature of liquid air is sufficient to develop marked electrical and luminous effects, which are again observed, when the crystal is taken out of the liquid, during its return to normal temperature. Since both liquid hydrogen and liquid air are good electrical insulators, the fact that electric discharges take place in them proves that the electric potential generated by the cooling must be very high. A crystal of nitrate of uranium indeed gets so highly charged electrically that it refuses to sink in liquid air, although its density is 2.8 times greater, but sticks to the side of the vacuum vessel, and requires for its displacement a distinct pull on the silk thread to which it is attached. Such a crystal quickly removes cloudiness from liquid air by attracting all the suspended particles to its surface, just as a fog is cleared out of air by electrification. It is interesting to observe that neither fused nitrate of uranium nor its solution in absolute alcohol shows any of the remarkable effects of the crystalline state on cooling.

_Cohesion._--The physical force known as cohesion is greatly increased by low temperatures. This fact is of much interest in connexion with two conflicting theories of matter. Lord Kelvin's view was that the forces that hold together the ultimate particles of bodies may be accounted for without assuming any other forces than that of gravitation, or any other law than the Newtonian. An opposite view is that the phenomena of cohesion, chemical union, &c., or the general phenomena of the aggregation of molecules, depend on the molecular vibrations as a physical cause (Tolver Preston, _Physics of the Ether_, p. 64). Hence at the zero of absolute temperature, this vibrating energy being in complete abeyance, the phenomena of cohesion should cease to exist and matter generally be reduced to an incoherent heap of "cosmic dust." This second view receives no support from experiment. Atmospheric air, for instance, frozen at the temperature of liquid hydrogen, is a hard solid, the strength of which gives no hint that with a further cooling of some 20 degrees it would crumble into powder. On the contrary, the lower the scale of temperature is descended, the more powerful become the forces which hold together the particles of matter. A spiral of fusible metal, which at ordinary temperatures cannot support the weight of an ounce without being straightened out, will, when cooled to the temperature of liquid oxygen, and so long as it remains in that cooled condition, support several pounds and vibrate like a steel spring. Similarly a bell of fusible metal at -182° C. gives a distinct metallic ring when struck. Balls of iron, lead, tin, ivory, &c., thus cooled, exhibit an increased rebound when dropped from a height; an indiarubber ball, on the other hand, becomes brittle, and is smashed to atoms by a very moderate fall. Tables XII. and XIII., which give the mean results of a large number of experiments, show the increased breaking stress gained by metals while they are cooled to the temperature of liquid oxygen.

TABLE XII.--_Breaking Stress in Pounds of Metallic Wires 0.098 inch in
diameter._

+15° C. -182° C.
Steel (soft) 420 700
Iron 320 670
Copper 200 300
Brass 310 440
German silver 470 600
Gold 255 340
Silver 330 420

TABLE XIII.--_Breaking Stress in Pounds of Cast Metallic Testpieces;
diameter of rod 0.2 inch._

+15° C. -182° C.
Tin 200 390
Lead 77 170
Zinc 35 26
Mercury 0 31
Bismuth 60 30
Antimony 61 30
Solder 300 645
Fusible, metal (Wood) 140 450

In the second series of experiments the test-pieces were 2 in. long and were all cast in the same mould. It will be noticed that in the cases of zinc, bismuth and antimony the results appear to be abnormal, but it may be pointed out that it is difficult to get uniform castings of crystalline bodies, and it is probable that by cooling such stresses are set up in some set of cleavage planes as to render rupture comparatively easy. In the case of strong steel springs the rigidity modulus does not appear to be greatly affected by cold, for although a number were examined, no measurable differences could be detected in their elongation under repeated additions of the same load. No quantitative experiments have been made on the cohesive properties of the metals at the temperature of boiling hydrogen (-252°), owing to the serious cost that would be involved. A lead wire cooled in liquid hydrogen did not become brittle, as it could be bent backwards and forwards in the liquid.