William Ramsay.

The gases of the atmosphere, the history of their discovery online

. (page 13 of 16)
Online LibraryWilliam RamsayThe gases of the atmosphere, the history of their discovery → online text (page 13 of 16)
Font size
QR-code for this ebook

The reasons for this belief were stated by Professor
Kamsay in an address given to the Chemical
Section of the British Association, at its meeting
at Toronto in 1897. As it may appear wonderful
that the existence of new and undiscovered
elements can be thus prophesied, an attempt will
be made to make clear the arguments in favour of
the forecast.

Not long after John Dal ton, in 1803, had re-
introduced the old Greek hypothesis of the atomic
constitution of matter, and had made his somewhat
unsuccessful attempt to determine the relative
weights of the atoms of the elements, speculation
began as to some possible relationship between the
weights of these atoms. These speculations finally,


as has already been remarked, culminated in the
periodic table, reproduced on p. 220. The last
column of that table contains the elements helium
and argon. The elements of preceding groups
show approximately regular differences between
their atomic weights ; thus, for example, the differ-
ence between the atomic weights of nitrogen, 14,
and phosphorus, 31, is 17; that between oxygen,
16, and sulphur, 32, is 16 ; hydrogen and fluorine
show a difference of 18, and fluorine and chlorine
of 16*5 ; and lithium, sodium, and potassium have
differences of 16 and 16*1 respectively. It was
highly probable, therefore, that an element should
exist, having an atomic weight about 16 units
higher than that of neon, and about 17 or 18
units lower than that of argon. It should have a
brilliant spectrum ; it should be a gas with a
boiling-point when liquefied higher than that of
helium, yet lower than that of argon ; like them
it should be monatomic, and it should display
inactivity in resisting combination with other
elements. Similar arguments would lead to the
conclusion th#t other two elements of higher
atomic weights should also be found one with
an atomic weight somewhat higher than that of


bromine, 80, but somewhat lower than that of
rubidium, 85*4 ; and that a third should succeed
iodine, with atomic weight greater than 127, but
less than 133. As no elements are known in
the chlorine or sodium column with still higher
atomic weights, it was imagined that it would be
unlikely that any element with a higher atomic
weight than, say, 130 would be discovered belong-
ing to the helium column.

But where were these elements to be sought ?
A very large number of minerals were heated in
a vacuum, and the gases they gave off extracted
by pumping ; some few yielded no gas whatever ;
but the majority evolved carbon monoxide and
dioxide, and hydrogen, in small quantity, while a
considerable number evolved helium, and one, a
mineral named malacone, containing zirconium,
evolved both helium and argon. The spectra of
the inactive gases were carefully examined, but
showed no unknown lines. The helium from
mineral waters, too, was introduced into vacuum-
tubes, but its spectrum likewise failed to show the
presence of any new constituent. The diffusion
of helium, which might have been expected to
separate a light from a heavy constituent of the


mixture, was also unsuccessful in revealing any
impurity, except a trace of argon ; the only clue,
and that not a very hopeful one, was that argon,
when systematically diffused, gave two portions
one slightly heavier, the other slightly lighter,
than the original gas. But the difference was
extremely minute, and was probably to have been
accounted for by experimental error.

However, as all other possible sources had been
examined, it appeared to be the only one left
untried ; and after an examination of sea- water,
which proved fruitless, a large quantity of argon
was separated *from the atmosphere, with the view
of its liquefaction and distillation a process which
would separate small quantities of light and heavy
constituents more perfectly than any other method.

The boiling - point of argon, at atmospheric
pressure, is 86*9 absolute, or -186*1 centigrade;
hence, in order to liquefy it, a plentiful supply of
liquid air was necessary. Dr. William Hampson,
who had devised an apparatus which yields liquid
air easily and in large quantity, kindly placed a
litre at the disposal of Professor Ramsay and his
coadjutor Dr. Travers. This liquid air was not
used, however, for the liquefaction of argon, but


experiments were made with it, to obtain practice
in manipulation, before risking the fifteen litres
of argon which were ready for liquefaction. On
the chance that the " dregs " or last residues of
this air might contain some one of the supposed
higher-boiling constituents of the atmosphere, the
final portions, after almost all had boiled away,
were collected ; the sample consisted largely of
oxygen, because, as nitrogen has a lower boiling-
point than oxygen, and is more volatile, the
spontaneous evaporation of the air had deprived
it of most of this constituent. On removal of the
oxygen by means of red-hot metallic copper, and
of nitrogen by magnesium, the inert residue was
examined spectroscopically. While it showed the
well-known argon spectrum, two brilliant lines
were also visible one in the yellow and the
other in the green part of the spectrum. The
density of the sample was 22 '47 ; hence it was
considerably higher than that of argon, which,
it will be remembered, is approximately 20. The
ratio of the specific heats of the sample was found
to be normal, viz. 1*66 ; hence this gas, like helium
and argon, is also monatomic. This gas was
discovered on May 30, 1898, and was named


" krypton," or hidden a name which had previously
been considered as a possible one for argon.

The fifteen litres of argon were next liquefied,
by causing it to enter a bulb, surrounded by liquid
air, boiling under reduced pressure. The liquid
argon, which occupied about 11 cubic centimetres,
was seen to be a colourless, mobile liquid ; it could
easily be frozen, by a slight reduction of tempera-
ture ; and it then formed a white, ice-like solid.

The first portions of the gas which boiled off
this liquid were collected separately, and examined
with a spectroscope ; a complicated and extremely
beautiful spectrum was observed, consisting of a
great number of red, orange, and yellow lines.
The density of this sample was 14*67. These
densities, of course, must not be taken as final
numbers, but merely as indicating that the argon
obtained by the evaporation of air contained a
heavier companion, and that the gas distilled
from liquid argon contained a lighter constituent,
than argon itself. To this constituent the name
" neon " or new was given.

The recognition of the presence of new gases in
the atmosphere, and their separation on a scale
sufficiently large for their study, are two very


different things. While the gases were discovered
in June 1898, it was not until October 1900 that
the investigation of their properties was completed.

It was first necessary to prepare them in
considerable quantity. And two quite distinct
operations were required to separate the lighter
constituent, neon, from air, and the heavier con-
stituent, krypton. Although both processes were
immediately commenced, as soon as a suitable
machine for producing liquid air had been pro-
cured from the Brin Oxygen Company, through
Dr. Hampson, it will conduce to clearness to
describe the operations separately.

The preparation of neon was carried out by
liquefying air, compressed to about 1-1 50th of
its volume ; this is achieved by allowing the com-
pressed air to pass downwards through a tightly-
wound copper coil, enclosed in a thin metal box,
which box is -itself surrounded by a packing of
wool, contained in an exterior case. The air
escapes at the bottom of the coil. Air, thus
compressed, approximates to a liquid, in so far
as its molecules or ultimate particles are so close
together that they attract one another. Now, in
order to separate particles of water from each


other, two methods are possible : either heat may
be applied, when the water changes to steam, a
body occupying a much larger volume than the
same weight of water; or if pressure is removed
from the surface of the water by means of a pump,
the particles at the surface fly off, while the tem-
perature of the water is lowered. With compressed
air, allowed to expand through a valve, a somewhat
similar phenomenon takes place. The air-particles,
separating suddenly from each other, absorb heat,
for the attraction between the particles is over-
come, and to effect this, heat is necessary ; as,
however, little external heat is allowed to enter,
the heat is derived from the air itself, during the
act of expansion ; and the air is cooled. The cold
air passes upwards over the copper coils through
which the compressed air is passing downwards ;
and the copper pipe is cooled. The effect naturally
is that the air passing downwards is progressively
cooled to a lower and lower temperature; and
finally, its temperature is so greatly reduced that
it issues in the state of liquid. The process is a
wonderfully rapid one ; in less than ten minutes
after the compression-pump starts, liquid air begins
to escape through the valve.


Of the two main constituents of air, nitrogen
has the lower boiling-point, 194*4, for oxygen
boils at 183. Hence, the liquefied portion of
air, which probably does not exceed one-twentieth
of what passes through the coils, contains a re-
latively larger proportion of oxygen than of
nitrogen, while the escaping portion consists more
largely of nitrogen. The connecting pipes were so
arranged that the issuing gas returned to the com-
pressor, to be again forced into the coils, and partly
liquefied. In this manner, the heavier constituents
of the air were condensed out, and the lighter
constituents, on compressing the remainder at a
pressure of about two atmospheres into a bulb
immersed in liquid air made to boil at a low
temperature (about 205), by being connected
with a vacuum-pump, were liquefied. This liquefied
portion, of course, contained the lower - boiling
constituents. Air was then blown through the
liquefied portion, causing it to evaporate; and
about one-third was collected in a large gas-holder.
The operation was repeated until a considerable
quantity had been obtained. It was then freed
from nitrogen by help of magnesium dust and
lime ; and the residue, consisting chiefly of argon,


but containing also neon, was liquefied and frac-
tionated by distillation.

And next began a tedious and troublesome
process of repeated fractionation ; the gas was
liquefied, distilled, and collected in successive
fractions in a methodical manner, until the
heavier and higher boiling argon had been sepa-
rated from lighter and lower boiling constituents.
Finally, a quantity of gas free from argon was
obtained, which would not liquefy, however much
the temperature of the liquid air which was used
as a cooling agent was lowered. This gas showed
the spectrum previously described as that of neon ;
but in addition, it showed the well-known lines
of helium. The spectrum of helium had previously
been recognised in that of atmospheric argon by
Drs. Kayser and Friedlander, but it was not
believed to contain helium in such considerable
amount. The helium appeared to form about
one-third of the total volume of gas ; the remain-
ing two -thirds were neon. As neither of these
gases can be liquefied at the temperature of
liquid air, even when its temperature is lowered
as far as possible by boiling it in a vacuum, it
was impracticable to separate them by distillation.


Many months were spent in attempting to effect
a separation by diffusion ; partial success was
attained, but perfect separation was impossible.
Dissolving the gases in liquid oxygen, and frac-
tionation from the solvent were also attempted, but
without success. Finally, an apparatus somewhat
similar in design to Dr. Hampson's air-liquefying
apparatus was constructed by Dr. Travers, in
which hydrogen was liquefied, and at the very low
temperature of boiling hydrogen, 20 '5 absolute,
the neon froze solid, while the helium remained
gaseous ; the helium was removed by help of a
pump, and the solid neon was allowed to warm
up, gasified, and collected separately. It may
be mentioned incidentally that, on removing the
liquid hydrogen from the bulb which contained
solid neon, the atmospheric air froze on it, and
encrusted it with a snowball of solid air, which
melted, dropped, and evaporated.

The krypton was prepared in a different
manner. It was left in the residue from about
thirty litres of liquid air, which had been used
for various operations. Instead of allowing the
last fractions of this air to boil away and mix
with the atmosphere, they were collected sepa-




rately and deprived of oxygen, which formed the
main bulk, and also of nitrogen, in the usual
manner. The residue consisted mainly of argon ;

FIG. 7.

and the lighter gases, helium and neon, had
already evaporated. This argon, which amounted
in bulk to several litres, was liquefied, and frac-
tionated in an apparatus of which an idea may
be gained from the annexed figure. The bulb (b)


was cooled by being immersed in liquid air ;
the argon was introduced from a reservoir (a)
through the stopcock, under some pressure, and
liquefied ; and the first portions were allowed to
boil back into the gas-holder. The remainder
was again liquefied, and separated into six

It was found, on inspecting the spectrum of
the lowest boiling of these fractions, that little
krypton was present, and that the gas consisted
mainly of argon. It is fortunate that the brilliant
yellow and green lines of krypton render its
recognition especially easy ; hence we had no
hesitation in rejecting gas in the spectrum of
which these lines were not visible ; but the
intensity of the spectrum misled us at first in
estimating the relative quantity of krypton in the
gaseous mixture.

The remaining fractions were re-fractionated
again and again ; but we found that anomalous
results were being obtained, for it appeared that
while krypton could be pumped away slowly,
even while the bulb was surrounded by liquid
air, a residue remained which, after removal of
the air-jacket, gasified. It turned out to be still


another gas, to which the name "xenon," or the
stranger, was given.

To describe in detail the numerous fractiona-
tions by which a separation of argon, krypton,
and xenon were effected would be tedious. The
final result was that two gases were obtained
which were not further altered by repeated frac-
tionation one of density 40*8, and the other of
density 64*0. These numbers imply the atomic
weights 81 '6 and 128, for reasons already given
on p. 197. The total amount of these gases was
disappointingly small ; for the krypton was only
about half a fluid ounce (about 15 cubic centi-
metres) in volume, while the xenon had barely
a quarter of that volume.

An attempt has been made to estimate ap-
proximately the amounts of these gases in atmo-
spheric air. The process for krypton and xenon
was to compress the ordinary atmospheric air,
and to run it through the liquefier ; the air which
had escaped condensation was passed through a
large gas-meter. During the experiment no less
than 1797 kilograms, or nearly 400 Ibs. of air,
passed the meter, or about 4500 cubic feet, while
11 '4 kilograms or 25 Ibs. were liquefied. The total


quantity was therefore 191 kilograms of air. It
may be remarked in passing that the liquid air
contained twice the normal quantity of argon, for
during liquefaction more argon liquefies proportion-
ally than nitrogen.

The liquid air was boiled down in a large glass
flask under a partial vacuum ; its boiling-point was
195 C. ; the residue measured about 200 cubic
centimetres. This liquid was then allowed to boil,
and the resulting gas was passed over red-hot
copper contained in a large tube ; much oxygen
was absorbed, and the final volume of gas was
50 litres ; after all nitrogen had been removed by
magnesium-lime mixture, the- argon left measured
12 '5 litres. It was then fractionally distilled, so as
to separate the argon, boiling at a low temperature,
from the krypton and xenon, of which the boiling-
points are much higher. Then the gases were
separated from each other by a long series of
operations of the same kind.

On comparing the volumes of krypton and
xenon with that of the air from which they had
been obtained, surprisingly small quantities were

The estimation of the helium and neon was


made in quite a different manner. Sir James
Dewar made the ingenious discovery that if air
be admitted into a vessel containing charcoal cooled
to a low temperature by liquid air surrounding
it, the oxygen and nitrogen are absorbed, while the
helium remains uncondensed, and not much neon
condenses. This property of gases to be absorbed
by charcoal has been long known ; but Dewar was
the first to apply it at low temperatures.

A measured quantity of air, about 17 litres,
was admitted to a bulb containing 100 grams of
cocoa-nut charcoal, cooled to 100 C. A large
proportion of the oxygen and nitrogen was
absorbed, but the neon and helium remained un-
absorbed. The unabsorbed gas was pumped off,
and again treated in the same manner with a
smaller quantity of charcoal, so as to remove
most of the oxygen and nitrogen. The remaining
mixture, which still contained a little nitrogen and
oxygen, was mixed with excess of oxygen, and
sparked over caustic soda. The residue of inert
gases was measured, after withdrawal of excess of
oxygen by phosphorus. The mixture of neon and
helium was then admitted to a bulb containing
charcoal cooled to 185 C., when the neon was


almost completely absorbed, the helium being left.
It was pumped off and measured, and when the
charcoal warmed up, the neon was also collected
and measured.

The amounts of these rare gases in crude argon
were found to be

Helium . 1 part in 2300 of argon by volume.

Neon . 1 757

Krypton . 1 200,000

Xenon . 1 1,700,000

In air, the gases are present in the following
proportions, approximately

Helium . . 1 part in 245,300 by volume.

Neon . . 1 80,800

Argon . 1 106-8

Krypton . . 1 ,,20 millions

Xenon . . 1 170 millions

It is surprising to think that there is much more
gold in an average sample of sea-water than there
is xenon in the air.

The density of pure argon, freed from these
gases, was determined by Sir William Eamsay and
Dr. Travers ; the two most reliable determinations
gave the figures 19*952 and 19*961. But knowing
the relative volume of neon and helium in crude


argon and their density, the density of pure argon
can be calculated ; it is 19*953 a number in close
agreement with the result of direct experiment.

To determine the properties of these elements,
apparatus had to be constructed on a very minute
scale. The boiling-points of argon, krypton, and
xenon were determined at all pressures between a
few millimetres and forty metres of mercury ; that
of argon at atmospheric pressure is 186*1 below
zero centigrade ; that of krypton, 1517 ; and that
of xenon, 109*1. On compressing these gases, the
highest temperature at which liquefaction occurs,
or the critical temperature, is for argon, 117*4,
for krypton, 62*5, and for xenon, -h 14*7 ; hence
on compressing xenon to about 50 atmospheres on
a cold day, it liquefies. They all form transparent,
colourless liquids, indistinguishable in appearance
from water ; and they likewise all freeze to white,
ice-like solids.

The position of the atmospheric elements in the
periodic table may be seen on p. 221. But it will
conduce to clearness if the elements are placed in
juxtaposition to those of neighbouring atomic
weights ; and an excerpt from the periodic table is
therefore introduced.













































Professors Lothar Meyer and Mendele'eff, after
constructing the periodic table of the elements,
drew attention to the fact that the physical
properties of the elements are periodic functions
of, or vary regularly with, the atomic weights of
the elements in each column : thus, for example,
we have in the first column of the elements, in
the table above, the elements hydrogen, fluorine,
chlorine, bromine, and iodine. For some unknown
reason, the first element of such a column diverges
considerably in properties from the rest ; thus there
is no great analogy between the behaviour of
hydrogen and that of the halogens, as the remain-


ing elements of the column are termed. But con-
trasting the latter among themselves, we see that
while the colour of fluorine is pale yellow, that
of chlorine is a darker greenish-yellow ; that of
bromine -gas red, and that of iodine-gas violet.
Again, fluorine has the lowest boiling-point, and
iodine the highest ; fluorine is the most active, or
in this column the most highly electro-negative
element of the group, and iodine the least ; for
fluorine attacks almost every other element, and
indeed compound, forming fluorides; while iodine
is comparatively inactive, and is somewhat difficult
to induce to combine with electro-negative elements.
On the other hand, while iodine forms relatively
stable compounds with that other highly electro-
negative element, oxygen, no compound of fluorine
and oxygen has been isolated.

Now in certain cases such properties admit of
a numerical value being attached to them. For
example, the volume of liquid or solid occupied by
a weight of the element taken in grams, at some
suitable temperature, implying the same condition
for each, is seen to vary progressively, with increase
in atomic weight. Thus, in the instance chosen
the volumes in the liquid state at the boiling-point




of 1 gram of hydrogen, of 19 grams of fluorine, of
35*5 grams of chlorine, of 80 grams of bromine,
and of 127 grams of iodine are

C.cs. .

Hydrogen. Fluorine. Chlorine. Bromine. Iodine.
14-3 17-15 23-5 27'1 34'2

In the annexed figure it will be seen that with
ordinates as atomic weights, and with abscissae as




10 -j

JO 2Q 30 40 50 60 70 80 90 100 110 120 130 140

Atomic Weights.
FIG. 8.

atomic volumes, broken curves are obtained both
in the fluorine and in the argon column of elements
which show such periodic variation. The same
kind of irregular variation of properties with atomic
weights is to be found if other properties be con-
sidered. Thus, when light passes through any
transparent material, it suffers more or less retarda-


tion, depending on the nature of the material. If
the retardation due to passage through a known
length of air at a certain temperature and pressure
is taken as unity, that due to passage through the
same length of the new gases under similar con-
ditions of temperature and pressure is shown

Helium. Neon. Argon. Krypton. Xenon.
0-124 0-235 0-968 1-450 2'364

Here, again, the increase with rise in atomic weight
is well defined. These numbers display among
themselves a very simple relationship, as has been
pointed out by Mr. Olive Cuthbertson. If the
value of the refractivity of helium be taken as
^, then the series becomes

Helium. Neon. Argon. Krypton. Xenon.
1 4 6 10

Mr. Cuthbertson has also shown that a similar
relationship holds for other elements which can be
obtained in the form of gas. Thus, if certain
columns of the periodic table (see p. 221) be

1 2 3 4 5 6 7 8 9 10 11 13 15 16

Online LibraryWilliam RamsayThe gases of the atmosphere, the history of their discovery → online text (page 13 of 16)