Images de page

Considerable uncertainty attaches to the speed tabulated for the spiral nebulae, as the radial velocities of comparatively few have been thus far determined; future work may change the value given, but it seems certain that it will remain very high. It will be seen at once that the high average speed of the planetary nebulae makes it difficult to adjust them to a position in the gamut of the stellar progression preceding that of the early type stars. If the planetaries are to be placed at this point, and if it be admitted that they are a primordial, but relatively short-lived, stage in stellar development, how is it possible to bridge over the great gap in spacevelocity which exists between them and the Class B stars? As matters stand, the planetaries have average space-velocities six times those of the Class B stars; nor does it seem reasonable to place them at the end of the stellar progression, after the Class M stars. It is true that the average velocity given for the Class B stars is that determined from the brighter members of the group, over one hundred times as bright as the brightest of the central stars in planetary nebulae, and a perfectly fair comparison must await that presumably very distant day when we shall have determined the radial velocities of a large number of faint stars. Some evidence is now being collected which may eventually show that relative mass is a contributing, perhaps even the main, factor in the function of space-velocity, rather than spectral type and relative stellar age. Should this theory supplant the present belief, it could scarcely remove the dilemma involved in postulating the planetary, of extremely rare occurrence, even with a space-velocity according more closely with that of the fainter stars, as a stage in stellar evolution through which the great majority of the stars are assumed to have passed. In view both of the high average space-velocity and the rare occurrence of the planetary nebulae, it would seem much more reasonable, as suggested by Dr. Campbell, to regard them as merely sporadic cases of stellar evolution, presumably of cataclysmic or collisional origin.

At this point the objection may very well be raised that the rarity of the planetary class is only apparent, and that large numbers of very small and very faint planetaries are still awaiting discovery. All the available evidence is against this point of view. The Harvard spectrographic survey, now nearly completed, will increase the number of classified stellar spectra from about 9000 to over 200,000. This increase of over two thousand per cent has produced but one new object with a spectrum of planetary type.

As a further check on this point I have examined seventy-nine small nebulae with the slitless spectroscope attached to the Crossley Reflector. This little instrument is exceedingly efficient in quickly detecting the gaseous character of the spectrum of a small or very faint object. The spectral image of a planetary nebula in this spectroscope will, in general, be seen only in the light of the main nebular line, appearing as a point, if stellar, and as a disk, if of larger size. All stars and small "white" nebulae are drawn out into lines, and an object of planetary type spectrum stands out at once in the field of view with extraordinary distinctness. The seventynine nebulae were chosen as far as possible within about 25° from the Milky Way, and were in general of the type described in the N.G.C. as small, very small, or stellar. Of those examined, seventy-eight gave a continuous spectrum, as shown in the following table:

[merged small][table]

But one object of the seventy-nine, N.G.C. 7139, was suspected to be of gaseous spectrum when observed with the slitless spectroscope, and subsequent photographs showed that it was of indubitable planetary form; the bright-line character of its spectrum has since been confirmed with a spectroscope attached to the 36-inch refractor (see Figure 74). The fact that but one object out of seventy-nine small nebulae has proved to be of planetary type would further support the view that no very great increase in the proportion of planetary nebulae is to be expected from future surveys.

March, 1918.










In Volume III, Publications of the Lick Observatory, Professor Keeler has described the efforts of several pioneers in stellar spectroscopy to determine the accurate wave-lengths and the chemical origins of the two strong green lines in the spectra of the bright-line nebulae. Keeler's contribution to the same problem, made with the 36-inch refracting telescope and the Brashear spectroscope, began, in effect, with measures of the displacements of the H/3 hydrogen line in the Orion nebula and in the planetary nebula N.G.C. 6572 (G.C. 4390, or 26, R. A. = 18" 7m), with reference to the corresponding laboratory line of hydrogen. The measures gave him the radial velocities of the two nebulae, and these enabled him to reduce the measured wavelengths of the two green nebular lines in the same spectra to their normal, or zero-velocity wavelengths. Knowing their normal wave-lengths, the observed wave-lengths of the same lines in the spectrum of any nebula could be converted into the observed radial velocity of that object. This logical sequence was used by Keeler as the basis for determining the radial velocities of twelve additional planetary nebulae, in 1890-91—altogether the radial velocities of the Orion nebula and thirteen planetary nebulae—a work of superb excellence and high value, in which one of the present writers was privileged to assist in a minor capacity during the summer months of 1890. We quote Keeler's summarized results as to the radial velocities:

1. The radial velocity of the Great Nebula in Orion is a recession of 17.7 km. per sec. This nebula is therefore substantially at rest with reference to the great stellar system, as far as the line-of-sight component is concerned.

2. The observed radial velocities of the thirteen planetary nebulae lie between 64.7 km. per sec. approach and 48.5 km. per sec. recession.

About a decade later Keeler's results for four or five of the nebulae were confirmed at several observatories, by the photographic method; but up to the year 1912 the remaining objects on his list had not been so observed, nor was the method applied to other nebulae, with the single exception of the Wolf-Rayet star-planetary nebula B.D. +30°3639, by Duncan.1

Concurrently with the inauguration and progress of the programme of determining stellar radial velocities with the D. O. Mills spectrograph—beginning in 1896—Mr. Campbell had hoped to find time to Undertake a systematic extension of our knowledge of nebular radial velocities, to include all bright-line nebulae whose spectra could be photographed with exposures of practicable lengths. The surprising discovery, early in 1910, that in general the radial

i Lick Obs. Bull., 6, 59, 1910. See page 150, following.


« PrécédentContinuer »