One hundred and sixty-nine presents were announced as having been received since the last meeting, including, amongst others :—

W. W. Bryant, History of Astronomy; S. W. Burnham, General Catalogue of Double Stars (2 vols.); and Sir G. H. Darwin, Scientific Papers, vol. i., presented by the authors; Galileo, Opere, Edizio nazionale, vol. iii. part 2 and vol. xix., presented by the Italian Government; Lady Huggins, Memoir of Agnes and Ellen Clerke, presented by the author; Oxford Astrographic Catalogue, vol. iii.; Observatoire de Paris, Atlas Photographique de la lune (Loewy and Puiseux), fasc. 9; and Perth Observatory, Western Australia, Catalogue of 420 standard stars, presented by the Observatories; Sir N. Lockyer, Report of the Eclipse Expedition to Majorca, presented by the Solar Physics Committee.

Astrographic Chart; 134 charts (Algiers, Bordeaux, Paris, and San Fernando), presented by the French Government; 23 charts, presented by the Tacubaya Observatory, Mexico; and 76 charts, presented by the Royal Observatory, Greenwich.

Three transparencies of the moon from negatives of M. Puiseux, presented by Mr Knobel; 9 transparencies of the northern Milky Way, presented by Professor Max Wolf.

Medal to commemorate the Benjamin Franklin bicentenary (bronze), presented by the American Philosophical Society.

Spectroscopic Observations of Cyanogen in the Solar Atmosphere and in Interplanetary Space. By H. F. Newall, M.A., F.R.S.

The presence of cyanogen in the atmosphere of the Sun seems to be indicated by the distinct appearance of the cyanogen absorption bands at wave-length 3883 in the solar spectrum. But, so far as I am aware, no definite observations have been directed to settle the possible doubt as to whether the cyanogen is confined to the solar atmosphere. The possible alternatives are that it may be in the Earth's atmosphere, or in space between the Sun and the Earth.

[Note. Nov. 11.-Professor Dyson has kindly called my attention to evidence which had escaped my memory and is conclusive as to the presence of cyanogen in the chromosphere. Sir N. Lockyer found CN bands in the flash spectrum of the chromosphere photographed in India in 1898 (Phil. Trans. R.S., vol. cxcvii. A., p. 202, and Mem. R.A.S., vol. liv., App. p. [52]), and Professor Dyson himself found them in his observations in Sumatra in 1901 (Phil. Trans., vol. ccvi. A., p. 438, and Mem. R.A.S. lvii., App. [36]). I have accordingly modified one or two of the statements in the paper.]

During the past summer I have made some observations in attempting to elucidate this point. The method adopted consisted in photographing side by side on one photographic plate two solar

spectra taken successively in the light coming from the east and from the west ends of the equatorial diameter of the Sun. Lines of truly solar origin should show relative displacement in the two spectra in consequence of the difference of velocity in the line of sight arising from the Sun's rotation. This is a form of the wellknown method of Cornu for distinguishing between lines of solar and lines of telluric origin.

In the result I find that my photographs show displacement of the cyanogen bands similar to that exhibited by neighbouring lines of iron and other vapours. There are, however, also evidences in the photographs that there is a trace of superposed cyanogen bands exhibiting displacements other than those attributable to the Sun's rotation. Thus the evidence so far accumulated shows that there is cyanogen in the Sun's atmosphere rotating with the Sun, and there is also cyanogen between the Earth's surface and the Sun showing spectroscopic displacements that may be attributed to independent motion.

Before passing on to give the evidence collected, it may be well to give some considerations which may be held to justify the inquiry as to the seat of cyanogen absorption. I recognise the highly speculative nature of some of them, but I beg leave to set them forth, in company with the statement of several newly observed facts, which go some way towards justifying speculation.

One of the most marked features of the spectrum of comets is the bright bands of cyanogen at 3883 first discovered by Sir W. Huggins. It has generally been held that this incandescent cyanogen and the other carbon compounds evidenced by cometary spectra are emitted from the comet's head, driven out under the influence of the Sun's radiation. On this view the observed fact that periodic comets seldom display brilliant tails receives the explanation that each successive return of a comet to perihelion reduces the quantity of vapour in the possession of the comet's nucleus; such vapours, then, are set free in interplanetary space and seem to be spread out, mainly in the plane of the comet's orbit, under the action of repulsive forces. The old view that the freed vapours are swept up by the gravitative attraction of the planets and the Sun is now being replaced by the modern view that radiation-pressure drives the larger aggregates of molecules or ions outwards from the Sun. Schwarzschild's work has shown that things as small as molecules cannot, as such, be repelled by radiation-pressure. It is generally held that molecules are of linear dimensions of the order of 10-8 cms., whilst the wave-length of green light is of the order 5 × 10-5 cms. And Schwarzschild has shown that things smaller than a tenth of the wave-length of the radiation are more attracted by gravitation to the Sun than repelled by the pressure of the Sun's radiation.

But it is held that a great deal of the display of comets' tails is due to illuminated dust as is shown both by the spectrum and also by the polariscopic phenomena observed by Prazoumowski and others. Here again the dimensions of particles giving the polarisa

tion effects afford some difficulties; but it might be held that the gases, which the spectroscope shows to be present in the tail, are carried out by the repelled dust. Arrhenius, however, as far as I understand his present views, is ready to regard the molecules of gases and vapours as grouped in aggregates large enough to bring them under the sway of radiation-pressure.

The work of Deslandres and Bernard on the spectrum of Daniel's comet 1907 d (C.R., cxlv. 445) shows that bands of some gaseous substances were visible in the tail at 45 minutes of arc from the nucleus in the middle of August; and I learn from a letter from Mr. Evershed that, at Kodaikanal, he and Mrs. Evershed have succeeded in photographing the same comet with a prismatic camera, with the following results:-"The hydrocarbon bands extend a long way up into the tail, whilst the cyanogen bands near 3883, although the most brilliant radiations in the comet spectrum, are concentrated round the head."

Now, both in the case of the absorption bands of cyanogen seen in the solar spectrum and in the emission bands of the same vapour seen in the cometary spectrum, a peculiar feature is that the ultraviolet bands at 3883 are the most marked, and that the blue and violet bands are far more feebly seen than under certain conditions attainable in the laboratory (Liveing and Dewar, Proc. R.S., xxx. 494). Some years ago I made a series of observations on the spectra of carbon compounds commonly present in vacuum tubes. My observations (unpublished) related chiefly to phenomena in electrodeless discharges brought about by Professor J. J. Thomson's electro-magnetic method (Phil. Mag., 1891, vol. xxxii. p. 321, and Proc. Roy. Inst., xiv. p. 243), wherein a tube or bulb of considerable diameter, ranging from about 12 mm. to 100 mm., is surrounded by a few coils of thick wire, the ends of which are connected with two Leyden jars or plate condensers. The discharge of such charged condensers is accompanied by exceedingly rapid electrical oscillations, which give rise to induced oscillations (accompanied by luminosity) in the rarefied gas in the enclosed tube. The method made it possible to produce brilliant luminosity in the gas at much lower pressures than those then attainable in ordinary vacuum tubes provided with electrodes. The pressures measured were between o'60 mm. and o'005 mm. of mercury.

I will refer here only to two of the results indicated by my observations. One was that as the pressure was diminished the bands of nitrogen and cyanogen, which I specially studied, became faint and disappeared at the red end of the spectrum, whilst those in the violet and ultra-violet were left distinct, or even intensified. I found no case which did not conform with the generalisation, that as pressure is diminished the maximum intensity of the banded spectra is, pushed towards the shorter wave-lengths the electrical conditions being kept constant. The observations of Hasselberg and of Deslandres support this view with respect to nitrogen; and it would not be difficult to devise a set of experiments to decide whether my observations can be reconciled with those of Nutting

(Astroph. Jour., xx. 131), which, so far as they go on these lines, seem to indicate an opposite conclusion for the much higher pressures between 1 mm. and 21 mm. At any rate my own (laboratory) experiments have led me to the belief that in the solar spectrum the evidence points to the conclusion that cyanogen is at low pressure. It is obvious that the pressure must be low in the case of a comet's tail extending several thousands (if not millions) of miles from the nucleus.

The other observation to which I now wish to refer was that the spectra of hydrocarbons and nitrocarbons were far the most easily elicited in such tubes and bulbs, and that the processes involved in the production of luminosity in the vapours could be started under conditions which proved instability, in the sense that the gas was luminous in regions of the vacuum bulb where the electromotive intensity (E.M.F. per cm.) was zero.

Are

This observation suggested an alternative to the usually accepted view of the origin of the cyanogen, etc., in cometary spectra, as follows:-Is it not possible that the hydrocarbons, nitrocarbons, etc., which seem to be evidenced by the spectra of all comets, are always present in circumsolar space, and rendered incandescent by some processes connected either with the motion of the solid parts (including dust) of the head of the comet through the vapours, or with the emission of some influence from the comet head? we to say that all comets, wherever they may come from in the universe, and whatever their main material may be, always bring with them the cyanogen and hydrocarbons which give them luminosity? Or is it not more rational to say that the spectra of all comets are approximately similar, because they always find the same vapours spread in their path as they approach the Sun, and can only elicit the spectra of these vapours?

If we adopt this latter view, then, apart from the phenomena of repulsion, we have only to explain the production of luminosity, and variations in the intensity of it. And out of the embarrassing wealth of suggestions connected with the development of the corpuscular theory of matter, there is no difficulty in picking one or two modes of influence to which incandescence might be attributed. I will not dwell on this aspect, but will call attention to another mode of producing luminosity-wherein there is no obvious sign of high temperature-I mean the luminosity of air when it rushes through a small crack in the glass walls of a vacuum tube, a phenomenon that must be familiar to most of those who have been engaged in experimental study of such tubes. Two or three times I have been able to recognise the bands of nitrogen in the spectrum of the glow in such a crack. Now, such luminosity cannot be peculiar to nitrogen. If it is legitimate to assume that hydrocarbons and cyanogen would behave similarly, it is not a long step to assume that it is as likely that luminosity will arise when small specks of cometary dust or solid particles are rushing through the gas (possibly in a sensitive state), as that luminosity should be produced when the gas rushes past the walls of a small crack into vacuum.

However the luminosity may be produced, the absence of brilliant display in periodic comets might, on the hypothesis of the constant existence of the vapours in interplanetary space, be attributed to the gradual change in the surface of the solid parts of the head of the comet, or to the exhaustion of the emitted influence.

My present purpose is, however, not to elaborate views about comets' tails, but simply to point out that if one accepts radiationpressure and the phenomena of the repulsion of comets' tails, one is driven to admit the existence of vapours and gases in circumsolar space, however difficult it may be to give account of their quantitative distribution in such space. The great extensions of the corona (as was pointed out by Huggins in his Bakerian Lecture, 1885), and the equally great extension of comets' tails, demand the admission that at any rate temporarily gases and vapours must exist in "free space.' How long the clearing process lasts, which depends on the conflict between the pressure of radiation and the gravitative attraction of the Sun and the planets, is another

matter.

It is difficult to devise crucial observations which would decide the points raised by these views. For it is to be expected that gases and vapours, if present at all, would be in such a turmoil of rotation round the Sun that the integrated effect of their absorption could hardly give defined lines or bands of absorption, except in the case of accidentally related motions in the vapours at different distances from the Sun. The best way to proceed seemed to be to make special observations of the cyanogen bands, and so learn whether there really are any peculiarities to be found in their behaviour. Professor Hale has pointed out (Astroph. Jour., xxv. 310) that the cyanogen fluting at 3883 is very decidedly weakened at the limb of the Sun relatively to the centre of the disk. Humphreys has shown (Astroph. Jour., xxvi. 28) that the lines of the cyanogen bands are not appreciably displaced in the spectrum of the electric arc, even at the highest pressures used by him (100 atmospheres).

The instrument used for the observations of the cyanogen bands is a new grating spectroscope made by the Cambridge Scientific Instrument Company, and recently mounted for solar observation at the Cambridge Observatory. It forms part of the installation which is in course of construction for solar work. The general form of the installation has been indicated in a previous paper communicated to the Society (M.N., lxvii. p. 161.) The observatory is indebted for this equipment to the munificence of the late Frank McClean, F.R.S.

The spectroscope is of Littrow type, i.e. one and the same tube and object-glass serve both as collimator and camera, the slit and photographic plate being slightly separated in the focal plane of the common object-glass. The lens has an aperture of 4 inches [101 mm.], and is of focal length 14 feet [4267 mm.]. A Rowland