ON THE SPECTROSCOPE AND ITS 
APPLICATIONS 
Vv. 
OAS YO analysis, then, teaches us this great fact, 
J that solids and liquids give out continuous spectra, and 
that vapours and gases give out discontinuous spectra ; that 
is to say, that we get bright lines in different parts of the 
Spectrum, instead of having an unbroken light all over the 
spectrum. I might vary this statement by stating broadly 
that the radiation or giving out of light by solids and 
liquids is a general one, and that the radiation or giving 
out of light by gases and vapours, instead of being 
general, is in the main a selective one. 
' The tubes, to which reference has already been made, 
put us then in complete possession of a point which has 
already been arrived at by two different lines of investi- 
gation. A few years ago Dr. Frankland, in investigating 
the spectrum of hydrogen, which, as you know, accord- 
ing to the statement I have just made, ought to give a dis- 
continuous spectrum, discovered that, when observing the 
spectrum under very great pressure, he got a white 
light, and a continuous spectrum. Afterwards Dr. An- 
drews, another fellow of the Royal Society, who was 
working at the theory of vapours and the theory of liquids 
from a perfectly different stand-point, and who never 
thought of using a spectroscope at all, arrived at the 
conclusion that it was quite possible that vapours might 
be so condensed in almost every case, that by crushing 
them together, so to speak, you might really arrive ata 
liquid form of the vapour which you moat choose to 
investigate. I hope you will not think that these high 
physical investigations are not practical enough. Let me 
remind you that we do not know what they may lead to. 
Not only did Dr. Frankland determine that very dense 
gases and very dense vapours gave continuous spectra, 
but in another research, in which I have had the honour 
of being associated with him, we have shown that the 
spectrum of a vapour or of a gas does very much more 
than tell us merely what the gas or vapour experimented 
upon is; it in fact tells us something of the physical 
condition of that gas or vapour, that is to say, whether 
it is very rare or whether it is very closely packed 
together—whether it exists under a low or a high 
pressure, 
tigation which has not only an immense application in 
every chemical experiment with which the spectroscope 
has to do, but it has its story to tell and its aid to give 
concerning every star that shines in the heavens. We 
may state generally that, beginning with any one element 
in its most rarefied condition, and then following its 
spectrum as the molecules come nearer together, so as at 
last to reach the solid form, we shall find that spectrum 
become more complicated as this approach takes place, 
until at last a vivid continuous spectrum is reached. 
Spectrum analysis, then, if it merely differentiated 
between gases, vapours, solids, and liquids, and between 
gases and vapours in different states of pressure, would 
really be a new chemistry altogether ; and I have no doubt 
that the time is not very far distant when, not only in the 
chemist’s laboratory, but in a great many applications of 
the physical sciences, the spectroscope will be considered 
as necessary, and will be almost as much used, as a che- 
mical balance, and the sooner that time comes the better. 
But not only are we able to differentiate between 
different bodies, but the most minute quantities of sub- 
Stances can be determined by this method of research. 
The thing seems so impossible, that you may, some of 
ou, feel inclined to doubt my assertion when I tell you, 
bor instance, that Kirchhoff and Bunsen have calculated 
that the 18-millionth part of a grain can be determined 
by the spectroscope in the case of sodium ; that is to say, 
if in anything which I choose to examine by means of my 
spectroscope the quantity of sodium present amounts only 
Very fortunately for us, this is an inves | 
NA TURE 345 
to the 18-millionth of a grain, the spectroscope is perfect] = 
competent to take up that minute quantity, and bring i, 
out into daylight, so as to be detected with certaintye 
This reaction of sodium is so delicate, that if we examin 
any flame, burning in air, we almost invariably find sodium 
in it, for every particle of dust is impregnated with a 
sodium salt, probably sodic chloride. This is not to be 
wondered at, as two-thirds of the earth’s surface is 
covered by sea, which contains a considerable amount of 
sodium salts, and the fine spray, which is continually 
caused by the dashing of the waves, evaporates and 
leaves minute specks of salt which are carried over 
the whole land, and make themselves visible in our 
spectroscopes. Take another instance. Lithium is a 
substance the knowledge of the existence of which as 
a common element we owe entirely to the spectroscope; 
the 6-millionth part of a grain of this can be detected. 
If we examine anything for lithium, and do not get 
the characteristic red line, we know that not even the 
6-millionth of a grain is present. Strontium, again, can be 
discovered if only a millionth part of a grain is present. 
So much for the great power of spectrum analysis in its 
physical applications, and its dealing with minute quanti- 
ties of the elements which we know already, and this of 
itself would be of enormous importance. 
But the spectroscope does not stop here; it discovers 
the known elements under conditions where detection 
seemed almost impossible, and in which the old chemistry 
was powerless to helpus. Letus take, again, for instance, 
lithium. Lithium was only known formerly to exist in 
four minerals ; it is now known, thanks to the spectro- 
scope, to exist almost everywhere. If we were to take 
the ash of a cigar and introduce it into a colourless gas 
flame and examine the colouration with the spectroscope, 
we should get a spectrum of lithium ; and if we analysed 
in the same way the ash of milk, or the ash of blood, or 
of grapes, tea, sugar, &c., we should also find it. Dr. 
Miller has shown that, in the Wheal Clifford mine 
8co lb. of this salt are given every 24 hours, though 
before the advent of spectrum analysis no lithium was 
known to exist there. It has also been found in meteoric 
stones, in the water of the Atlantic, &c. Surely this is 
an application of very great importance. 
Another extremely important point about spectro- 
scopic analysis is that, although we may have to analyse 
a complicated mixture of substances, the spectroscope is 
perfectly competent to deal with them. The characteris- 
tic lines for each element must stand out and be visible 
whether the substance be simple or complex. Thus, for 
instance, if we mix together some sodium and lithium, 
and place some of the mixture in a flame, we shall see 
nothing but the brilliant yellow colour due to sodium, the 
crimson flame of the lithium being entirely hidden, A 
moment’s examination with the spectroscope, however, is 
sufficient to show us that both lithium and sodium are, 
without the slightest doubt, present in the flame ; for both 
the yellow and red lines stand out as distinctly as they 
did when the simple salts were experimented with. The 
presence of lithium, indeed, may be detected, even if it 
be mixed with ten thousand times its bulk of sodium 
compounds. ‘ ; 
But, further, spectrum analysis is not satisfied with 
showing us sources of known elements. It discovers new 
elements altogether. In 1860, Bunsen happened to be 
examining with a spectroscope the result of one of 
his analyses of the waters of a spring near Diirk- 
heim, and he saw some lines which he had never 
seen before, although he had very carefully mapped 
the spectra of the known elements. Bunsen, as you 
know, is a very resolute chemist, and what he did 
was this. Having faith in his instrument, he evaporated 
no less than forty-four tons of the water of this spring, 
and out of these forty-four tons he got about two hundred 
grains of what turned out to be a new metal, which he 
