[ocr errors]

ART. IV.-1. Spectrum Analysis. Six Lectures delivered in

1868. By HENRY E. Roscoe, F.R.S. 8vo. London: 1869. 2. Le Stelle: Saggio di Astronomia Siderale. Del P. A.

SecchI. Milano : 1878. 3. Researches in Spectrum Analysis in connexion with the

Spectrum of the Sun. By J. NORMAN LOCKYER, F.R.S.

• Proceedings of the Royal Society,' vol. xxviii. : 1879. 4. On the Spectra of some of the Fired Stars. By WILLIAM

HUGGINS, F.R.A.S., and W. A. MILLER, M.D., LL.D. Philosophical Transactions of the Royal Society,' vol.

cliv.: 1864. 5. Further Investigations on the Spectra of some of the Stars

and Nebulæ, with an Attempt to determine therefrom whether these Bodies are moving towards or from the Earth. By WILLIAM HUGGINS, F.R.S. Philosophical Transactions

of the Royal Society,' vol. clviii.: 1868. 6. The Universe of Stars. By RICHARD A. PROCTOR.

Second Edition. London: 1878. Wh THEN Kirchhoff demonstrated, twenty-one years ago, the

existence of sodium in the atmosphere of the sun, he made an advance of which we are even yet hardly in a position to estimate the full importance. The discovery supplied one more proof of the harmony of nature and the fundamental unity of science. The corruptible' materials of our motherearth were shown by conclusive evidence to form part of the incorruptible' substance of the radiant orbs of heaven. Astronomy, which had hitherto taken cognisance of matter only in its most general form, was now compelled to descend into the laboratory in order to study its various kinds and qualities, together with their mutual actions and relations. The science of celestial mechanics became, all at once, the science of celestial chemistry. From the new point of departure thus unexpectedly provided, untried fields of research were gradually perceived to stretch farther and farther away into the illimitable distance. The invention of the telescope does not, indeed, form a more noteworthy epoch in the history of astronomy than does the application of the prism to the physical investigation of the heavenly bodies. Yet, marvellous as are the results already achieved by spectrum analysis, they are as nothing compared with the crowd of unsolved problems which continually stimulate the curiosity, and baffle the skill,


of the spectroscopist. Nor should this occasion surprise. Since creation is modelled on a scale utterly incommensurable with human faculties, the progress of science necessarily proposes more questions than it answers, and opens up, one after the other, vistas of the unknown, each forming, as it were, a separate pathway towards the one infinity.

Thus each new discovery, by revealing previously unsuspected ignorance, suggests fresh efforts, and promotes fresh advances. Already the more hopeful among men of science look forward with confidence to the recognition of a law, higher and wider than that of Newton, embracing all the operations upon matter of the so-called physical forces,' and reducing under a common denomination the actions of gravity and cohesion, the phenomena of light, heat, and electricity. We venture indeed to assert that no one who earnestly and intelligently looks nature in the face can escape the conviction that such a principle regulates the apparent anomalies, and. harmonises the apparent contradictions, visible in the world around us. The generalisation of knowledge, however, becomes increasingly difficult with its extension; by the accumulation of particulars induction is rendered more sure, but is also rendered more arduous; and science is impeded in its progress in proportion as it is amplified in its details. We may then have to wait long for the realisation of the hopes held out to us, and must for the present content ourselves with noting effects where we would willingly penetrate into

Nevertheless, the close relationship more and more clearly perceived to unite the physical sciences forms in itself a species of generalisation, and will doubtless contribute in the future to maintain and increase the high intellectual importance of natural investigations.

The discovery of spectrum analysis has most markedly emphasised this relationship. The sciences of astronomy and chemistry can no longer be said to exist independently one of the other. The astronomer demands from the chemist an interpretation of what he observes ; the chemist turns to the astronomer for confirmation of what he divines. The working of this new alliance is strikingly exemplified in Mr. Norman Lockyer's recent investigations into the nature of the chemical elements. The sixty-five to seventy * different substances at


[ocr errors]

The exact number cannot at present be determined. Since 1877, claims have been put forward to the discovery of no less than fourteen new metals, in many cases, probably, on very insufficient grounds. See a paper in “Nature,' July 8, 1880.

[ocr errors]
[ocr errors]

present known to enter into the composition of the earth have long been regarded by chemists as only provisionally elementary in their character. The term "element' was simply meant to convey that hitherto they had not been decomposed; but it was clearly foreseen that with improved laboratory appliances many such bodies would be reduced to a simpler condition--a prevision already verified in the case of the allied substances, chlorine, bromine, and iodine.

But theory has, in this direction, far outstripped experiment. Between the atomic weights of the various elements, numerical relations, as remarkable as those connecting the different members of the solar system, have been perceived to exist. Empirical laws, of similar character to · Bode's law' of planetary distances, regulate the combining proportions of certain groups of substances analogous in their qualities, indicating, it is argued, varying degrees of complexity in their composition.* These ingenious speculations have even been made the basis of successful prediction. A gap in the series indicated by his periodic law' enabled Professor Mendelejeff, in 1869, to announce the existence and describe the qualities of a new metal, discovered, six years later, by M. Lecoq de Boisbaudran in a blende from the mines of Pierrefitte, and named by him 'gallium.'t Moreover, the striking fact that nearly all atomic weights are simple multiples of the weight of the hydrogen-atom gave rise to Prout's celebrated hypothesis of a primordial substance no other than hydrogen. But even this is not enough. Still bolder speculators derive from luminiferous ether—the refuge and the reproach of scienceevery form of ponderable matter; and the remarkable theory of vortex-atoms,' elaborated from profound mathematical considerations by Sir William Thomson and the late Professor Clerk Maxwell, has lent plausibility (it would be going too far to say probability) to what seemed at first sight an extravagant conjecture.

We learn then, without surprise, from a paper communicated to the Royal Society, December 12, 1878, the title of which we have prefixed to this article, that Mr. Lockyer has ·been led by his spectroscopic studies to doubt the elementary character of some, if not all, of those bodies which have hitherto successfully maintained that reputation. We are not prepared to deny his conclusions; but we venture to dissent from some at least of the arguments by which he seeks to support them.

* Chemical News, vol. xxxviii. p. 66.
† Comptes Rendus, t. lxxxi. pp. 493, 969.

His observations are of the highest interest and importance; but they seem to us hardly to warrant the interpretation which he puts upon them.

We need not here dwell upon the first principles of spectrum analysis; they were ably expounded in the pages of this Journal * shortly after their discovery, and are dwelt

upon with still fuller detail in the valuable work by Professor Roscoe which stands at the head of this article. It may, however, be well to remind our readers that while further enquiry has amply confirmed the fundamental theorem upon which the science rests-namely, that vapours absorb rays of the same refrangibilities that they radiate-a multitude of secondary facts have been recognised, which, although they at present tend somewhat to embarrass our conclusions, will no doubt eventually contribute to define them. Thus, while it may be looked upon as established that an incandescent solid or liquid body gives a continuous spectrum—in other words, emits light of every shade of colour—the converse no longer holds good. A continuous spectrum is not necessarily due to a solid or liquid, but may be derived from a vapour at considerable pressure. Many physicists, indeed, believe that the vast mass of the sun consists of glowing and enormously compressed gases, the fine black lines which rule the rainbow-tinted ribbon unrolled out of its light by the prism, owing their origin to the selective absorption of the same vapours at a higher level and reduced temperature and density. They in fact stop on their passage the identical rays that they more feebly emit, thus producing those innumerable minute gaps of comparative darkness in the sun's light known as “ Fraunhofer lines. Now each of these vapours or gases gives forth, when heated to incandescence, a more or less numerous set of lightwaves, strictly definite in their respective colours and consequent positions in the spectrum ; and it is by the identification of these beams, or bright lines, with corresponding dark lines in the solar spectrum, that inferences, surprising but entirely trustworthy, have been drawn regarding the physical constitution of our great luminary.

The spectroscopic evidence adduced by Mr. Lockyer in proof of the compound character of the elements' may be conveniently divided into three classes—terrestrial, solar, and stellar. His position would, indeed, be much more clearly intelligible if founded on some settled theory of luminous radiation by matter in its various conditions. But on this


* Edinburgh Review, vol. cxvi., art. “Solar Chemistry.'

point modern science has nothing to offer beyond some vague and unsatisfactory conjectures.

We find ourselves, at the very threshold of enquiry, confronted by the (at present) inscrutable relations subsisting between that enigmatical substance whose vibrations are light, and the gross matter originating those vibrations by its movements. This much, however, we may safely say. A vibrating molecule is, speaking generally, not a simple body, but a system, probably of extremely complex constitution. Now any disturbance affecting that system will be faithfully reflected in the rays of light, which are the visible translation of its intimate thrillings. Such disturbances may be almost infinitely various in kind and degree, the actual severance of the atoms, or parts constituting a molecule, being only one extreme case amongst a multitude of possible modifications. But, while it is certain that each infinitesimal variation of molecular relations must produce a corresponding effect in the spectrum, this severance of atoms, or dissociation,' is adopted by Mr. Lockyer as a general rationale of all spectroscopic changes.

His main argument under this head is founded on analogy. He observes that the spectra of bodies supposed to be simple undergo, in like circumstances, changes precisely similar to those of bodies known to be compound. In the latter case the explanation is obvious and undeniable. The spectrum characteristic of the compound gives place, as the temperature rises and dissociation proceeds, to the spectrum characteristic of its principal constituent. The easy and natural course seems to be to transfer this explanation to the other case. And it is this which Mr. Lockyer has taken. Now we are far from denying that chemical separations play a certain part in producing the appearances revealed by the prism ; what we contend is that the cause in question, far from being universally active, is most likely only exceptionally so. In the first place, marked changes occur in cases where there can be no question of dissociation. By the mere condensation or rarefaction of an incandescent vapour, the bright lines of which its spectrum is composed can be increased or diminished at pleasure. If we suppose, according to the received theory, the light-producing vibrations of minute particles of matter to be maintained by the mutual impacts of those particles, then the fewer the impacts, the more feeble the vibrations. And, just as the harmonics * of a musical note can

For an account of Mr. Johnstone Stoney's ingenious barmonic theory of line spectra, see Schellen's “Spectrum Analysis,' Appendix A (English translation).

« ForrigeFortsett »