514 
markings and the variations of the polar cap with the seasons. 
M. Fm. Touchet contributes an illustrated account of his suc- 
cessful attempt to photograph the ‘‘ shadow cast by the planet 
Venus.”’ This he did on January 11, with an exposure of fifteen 
minutes, the object casting the shadow being an incandescent 
bulb-holder placed about 21 cm. from the plate.—Lastly, rather 
more than nine pages are devoted to a dissertation, by M. 
Rideau, on ‘‘the satellites of Jupiter,” dealing with their 
dimensions, surface, probable variability of brightness, eclipse 
and other phenomena. 
SOLID HYDROGEN. 
[X the autumn of 1898, after the production of liquid hydrogen 
was possible on a scale of one or two hundred c.c., its 
solidification was attempted under reduced pressure. At this 
time, to make the isolation of the hydrogen as effective as 
possible, the hydrogen was placed in a small vacuum test-tube, 
placed in a larger vessel of the same kind. Excess of the 
hydrogen partly filled the circular space between the two 
vacuum vessels. The apparatus is shown in Fig 1. In this 
way the evaporation was mainly thrown on the liquid hydrogen 
in the annular space between the tubes. In this arrangement 
the outside surface of the smaller tube was kept at the same 
temperature as the inside, so that the liquid hydrogen for 
the time was effectually guarded from influx of heat. With 
such a combination the liquid hydrogen was evaporated under 
some IO mm. pressure, yet no solidification took place. Seeing 
experiments of this kind required a large supply of the liquid, 
other problems were attacked, and any attempts in the direction 
Fiel, 
Fic 2. 
of producing the solid for the time abandoned. During the course 
of the present year many varieties of electric resistance thermo- 
meters have been under observation, and with some of these 
the reduction of temperature brought about by exhaustion was 
investigated. Thermometers constructed of platinum and 
platinum-rhodium (alloy) were only lowered 14° C. by exhaus- 
tion of the liquid hydrogen, and they all gave a boiling point of 
—245°C., whereas the reduction in temperature by evaporation 
in vacuo ought to be 5° C., and the true boiling point from 
—252° to —253° C. In the course of these experiments it 
was noted that almost invariably there was a slight leak of 
air, which became apparent by its being frozen into an air snow 
in the interior of the vessel, where it met the cold vapour of 
hydrogen coming off. When conducting wires covered with 
silk have to pass through india-rubber corks it is very difficult 
at these excessively low temperatures to prevent leaks, when 
corks get as hard as a stone, and cementscrack in all directions. 
The effect of this slight air leak on the liquid hydrogen when the 
pressure got reduced below 60 mm. was very remarkable, as it 
suddenly solidified into a white froth-like mass like frozen foam. 
My first impressions were that this body was a sponge of solid 
air containing the liquid hydrogen, just like ordinary air, which 
is a magma of solid nitrogen containing liquid oxygen. The 
1 Read before the British Association (Section B), Dover Meeting, by 
Prof. James Dewar, F.R.S. 
NO. 1560, VOL. 60] 
NALORE 
[SEPTEMBER 21, 1899 
fact, however, that this white solid froth evaporated completely 
at the low pressure without leaving any substantial amount of 
solid air led to the conclusion that the body after all must be 
solid hydrogen. This surmise was confirmed by observing that 
if the pressure, and therefore the temperature, of the hydrogen 
was allowed to rise, the solid melted when the pressure reached 
about 55 mm. The failure of the early experiment must then 
have been due to supercooling of the liquid, which is prevented 
in this case by contact with metallic wires and traces of solid 
air. To settle the matter definitely, the following experiment 
was arranged. A flask, Cc, of about a litre capacity, to which a 
long glass tube bent twice at right angles was sealed, as shown 
in Fig. 2, and to which a small mercury manometer can be 
sealed, was filled with pure dry hydrogen and sealed off. The 
lower portion, A B, of this tube was calibrated. It was surrounded 
with liquid hydrogen placed in a vacuum vessel arranged for 
exhaustion. As soon as the pressure got well reduced below 
that of the atmosphere, perfectly clear liquid hydrogen 
began to collect in the tube 4 B, and could be observed accumu- 
lating until, about 30 to 40 mm. pressure, the liquid hydrogen 
surrounding the outside of the tube suddenly passed into a solid 
while foam-like mass, almost filling the whole space. As it was 
not possible to see the condition of the hydrogen in the interior 
of the tube 4B when it was covered with a large quantity of this 
solid, the whole apparatus was turned upside down in order to 
see whether any liquid would run down AB into the flask c. 
Liquid did not flow down the tube, so the liquid hydrogen with 
which the tube was partly filled must have solidified. By 
placing a strong light on the side of the vacuum test-tube 
opposite the eye, and maintaining the exhaustion to about 
25 mm., gradually the solid became less opaque, and the 
material in A B was seen to be a transparent ice in the lower part, 
but the surface looked frothy. This fact prevented the solid 
density from being determined, but the maximum fluid density 
has been approximately ascertained. This was found to be 
0086, the liquid at its boiling point having the density 0°07. 
The solid hydrogen melts when the pressure of the saturated 
vapour reaches about 55mm. In order to determine the tem- 
perature, two constant volume hydrogen thermometers were 
used. One at o° C. contained hydrogen under a pressure of 
269°8 mm., and the other under a pressure of 127 mm. The 
mean temperature of the solid was found to be 16° absolute 
under a pressure of 35 mm. All the attempts made to get an 
accurate electric resistance thermometer for such low tempera- 
ture observations have been so far unsatisfactory. Now that 
pure helium is definitely proved to be more volatile than 
hydrogen, this body, afier passing through a spiral glass tube 
immersed in liquid hydrogen to separate all other gases, must 
be compared with the hydrogen thermometer. For the present 
the boiling point which is 21° absolute at 760 mm., compared 
with the boiling point at 35 mm., or 16° absolute, enables the 
following approximate formula for the vapour tension of liquid 
hydrogen below one atmosphere pressure to be derived :— 
log £— 67341 — 83'28/T mm., 
where T = absolute temperature, and the pressure is in mm. 
This formula gives us for 55 mm. a temperature of 16°7° absolute. 
The melting point of hydrogen must therefore be about 16° or 
17° absolute. It has to be noted that the pressure in the con- 
stant volume hydrogen thermometer, used to determine the tem- 
perature of solid hydrogen boiling under 35 mm., had been so 
far reduced that the measurements were made under from one- 
half to one-fourth the saturation pressure for the temperature. 
When the same thermometers were used to determine the boil- 
ing point of hydrogen at atmospheric pressure, the internal gas 
pressure was only reduced to one-thirteenth the saturation pres- 
sure for the temperatures. The absolute accuracy of the boiling 
points under diminished pressure must be examined in some 
future paper. The practical limit of temperature we can com- 
mand by the evaporation of solid hydrogen is from 14° to 15” 
absolute. In passing it may be noted that the critical temper- 
ature of hydrogen being 30° to 32° absolute, the melting point 
is about half the critical temperature. The melting point of 
nitrogen is also about half its critical temperature. The foam- 
like appearance of the solid when produced in an ordinary 
vacuum is due to the small density of the liquid, and the fact 
that rapid ebullition is substantially taking place in the whole 
mass of liquid. The last doubt as to the possibility of solid 
hydrogen having a metallic character has been removed, and 
for the future hydrogen must be classed among the non-metallic 
elements. 
Ee 
