« PrécédentContinuer »
The Wave-Lengths of the Nebular Lines.—The most extensive observations of nebular wavelengths heretofore available have been those of Campbell,31 who investigated the spectra of the Orion and six planetary nebulae. Measurements of the spectrum of the Orion nebula have also been published by Lockyer,32 and two bright planetaries have been the subjects of investigation by Max Wolf.33 These observers measured a comparatively large number of lines with the use of low dispersion. Accurate observations have been made of a few of the brighter lines. Of these the earliest are those of Keeler,34 who in 1891 determined by visual methods, and with a high degree of precision, the wave-lengths of the two green lines now generally designated as N, and N-. These lines have been remeasured photographically by Hartmann,35 Campbell and Moore,36 and the writer.
A very beautiful application of the methods of interference measurement to an astronomical problem was made by Buisson, Fabry, and Bourget in their observations of the doublet near 3727A in the spectrum of the Orion nebula.37 They not only determined the positions of these lines with the greatest accuracy, but detected differences of wave-length in various parts of the nebida due to differential radial velocity.
The present determinations depend upon the measurement of the spectra of nine bright nebulae. The results are given in table 11. The table includes, with a few exceptions, all of the well authenticated lines announced by previous observers and a number in addition to these. Two of the exceptions are the very faint green lines at 518 and 531^, discovered respectively by Vogel and Campbell. They lie in a region of the spectrum to which it is difficult to sensitize plates and do not appear on any of my photographs. Should it ever become necessary to determine their places photographically plates can probably be specially sensitized for the green which will record them in a reasonable exposure time. Besides these lines the only others not measured are two reported by Max Wolf at 3629 and 3652A in the spectrum of N.G.C. 6543. There is a reference to them in the description of the spectrum of that object.
While quite a variety of spectrographs have been used, each observation being accomplished with the instrument which seemed best adapted to the particular purpose, most of the spectrograms have been made with the single-prism spectrograph (a) attached to the 36-inch refractor. Some of the brighter lines were, however, photographed with a battery of three prisms and with a grating, while in the ultra-violet it was necessary to depend very largely on the slitless quartz spectrograph. I regret that it has not been practicable to secure satisfactory photographs of all of the ultra-violet lines with the quartz slit spectrograph. The reason for the failure lies in the strong curvature of the field given by the simple quartz lenses. This curvature corresponds to a radius of six inches, and is too great to be accommodated by bending, even with the most flexible glass plates procurable. The use of films is unsatisfactory on account of certain changes in length which they have been found to undergo even in the comparatively short interval of time required for an exposure. It is not uncommon for a film to alter its dimensions by a half of 1 per cent in less than two hours, a degree of instability not to be tolerated in any photographic operation where accuracy is required. We have been attempting for several years to secure achromatic lenses which will give a better field and thus allow the use of glass plates, but disturbed commercial conditions have prevented the accomplishment of this purpose.
The instruments used are all described in the appendix under the head of the letters by which they are indicated in the text.
si Astron. and Astrophys., 13, 384 and 494, 1894. »» Ap. Jour., 15, 291, 1902.
as Phil. Trans., 186A, 73, 1894. 3° Lick Obs. Bull.. 9, 6, 1915.
33 Sits. Heidel. Akad, Wiss.. 35 Ab., 1911. « Ap. Jour., 40, 241, 1914. 3< Publ. Lick Obs., 3, 165, 1894.
It is believed that the observations are practically free from systematic error, a fact which is indicated by the agreement with lines of known origin where they are available for comparison. Certain of these known lines, particularly the brighter ones of the hydrogen series, have been used to reduce the nebular wave-lengths to normal, and a fair agreement with laboratory values follows as a necessity; but the fainter lines above He have, as a rule, not been used for this purpose, and here the degree of correspondence between the nebular and laboratory values may be taken as an indication of the accuracy of the measures. In the near ultra-violet there is a check furnished by the interferometer determinations of Buisson, Fabry, and Bourget of the wave-length of the doublet at 3727A. The comparison appears to indicate a systematic error at this point of +.06A, which I should hardly have expected. This quantity is, however, within the limits of accuracy of the measurements of any other lines in the vicinity, and is therefore a matter of little consequence. The error does not persist below He.
The observations published in the previous paper40 are included, with the more recent ones, in table 11.
Wave-lengths of the nucleus bands are collected in table 12. In the case of B.D. +30° 3639 and a few other objects the measurements were made chiefly with a slit spectrograph and the positions may be regarded as reliable. In most instances, however, the wave-lengths have been estimated on slitless spectrograms and are therefore subject to considerable uncertainty. Where a band is very broad and apparently composite the wave-lengths of the probable components are inserted and covered by a brace. For further particulars the observations recorded in Chapter I should be referred to.
The photographs on which these wave-lengths depend were made on Seed No. 27 and Seed No. 30 plates, for the photographic region, and for the "visible" region of the spectrum, on Seed No. 27 plates bathed in Pinacyanol, Pinaverdol, and Homocol in the proportions recommended by Wallace39.
Intensities of the Nebular Lines.—There are two systems of intensity used in this discussion, one of which refers to the intensity as photographed with the glass spectroscope (a) attached to the 36-inch refractor with correcting lens. This scale is designated by Ig. The other indicates the intensity as observed with the slitless quartz spectrograph used in conjunction with the Crossley reflecting telescope. This has been termed the Iq scale. It might appear at first that the use of the two scales would lead to confusion, and that the glass system should be discarded in favor of that determined by quartz, which is of course the more uniform of the two. In view, however, of the fact that a great deal of work in this field still lies ahead of the glass spectrograph, it seems better to preserve the older scale for the convenience of observers who may desire to compare observations made with this, or similar apparatus, with the present ones.
Ig System.—In the earlier paper40 the intensities of the bright nebular lines were indicated on an arbitrary scale running from 1 to 80, 1 indicating the faintest line photographed. In extending the work to the fainter lines all of these naturally fell within the unit, and desiring not to alter the original intensity assignments fundamentally, and at the same time to avoid the use of decimals, the intensities were all multiplied by ten, which allowed plenty of latitude for the estimates of the fainter lines. This scale as originally adopted was, as has been said, quite arbitrary, and was described as follows: "The relative brightness of the lines as photographed is indicated in a way by the numbers opposite the wave-lengths. About all that can be said of the system is that, for any one object, the larger the number the brighter the line. Such estimates must of course be used with caution, as the apparent brightness of a line is greatly influenced by photographic and instrumental conditions." These considerations apply to the present extension of the scale to the fainter lines.
3» Lick Obs. Bull., 5, 151, 1909. io Lick Obs. Bull, 1, 153, 1902.
and that the nebular line is actually Ds. This measurement has not been used in striking
values of N, and N, are the means, struck by Campbell and Moore, of the results of several O. Bull., 9, 9, 1915.:
•lose to the strongest line in the secondary spectrum of hydrogen observed by Dufour. It nly case of coincidence with a line of that spectrum, however, and too great significance should 1 to this single correspondence in wavelength.