A small very brilliant planetary and one that has been studied extensively by spectroscopic observers. There is a faint nucleus which, according to Keeler, who made a careful study of the telescopic appearance of the object as well as of its spectrum, is not stellar, but apparently due to a general increase in the density of the nebulous material. Reproductions of the spectra are given in figures 3 of plate XLIII and 3 of plate XLVII.

Reference has already been made to the fact that several of the bands characteristic of the spectra of the Wolf-Rayet stars occur in the spectrum of the nucleus. The line 4686A does not occur in the outer regions of the nebula, but is localized in the spectrum of the nucleus as a band

about 15 Angstroms wide. Other broad bands similarly restricted are 4057 and 5807A, while the lines 4634, 4641, 4650, and 4658A are hazy and also confined to the nucleus. Most of these bands are Wolf-Rayet radiations, and it was their detection in the nucleus of this object which led to a fuller investigation of the spectra of nebular nuclei.

In addition to those mentioned above, there are other short lines scattered throughout the spectrum. By a short line is meant one which does not extend beyond, or very far beyond, the spectrum of the nucleus. In a nebula, like this one, in which the intensity falls off gradually in passing from the center outward it is difficult to judge of the length of a line, since by making the photographic exposure long enough one can lengthen any line unless it is entirely absent from the outer parts of the nebula. Nevertheless the question of the relative lengths of lines is of such interest that I have, with a realization of these difficulties, ventured to describe many of the lines as short, medium, and long. Such a classification must, in the nature of things, be tentative, but I am inclined to believe, on account of the distinctive appearance of many of the lines, that the assignments are, on the whole, not far from correct. The estimates are given in table 6.

In his comments on the spectrum of this nebula Professor Wolf states that the longest line is Hy, which he finds to extend 27′′ from the center. It seems probable that his result is affected by instrumental irregularities, or unfavorable observing conditions, since the radius of the nebula is only 5" or 6".10 It is therefore doubtful whether this observation can be regarded as sufficiently delicate to settle the somewhat difficult question of the relative lengths of the lines in this nebula.

The wave-lengths and intensities of the lines are given in tables 11 and 13.

In addition to the emission bands already mentioned, the nucleus gives a continuous spectrum. As photographed with the slitless spectrograph this spectrum fades gradually toward the violet until the image at 3727A is reached. On the less refrangible side of the image it appears to increase in intensity, but that is, I believe, an effect due to the crowding together of the images formed by the hydrogen series, which are very close together. There is a further apparent increase at about 3650A, which is undoubtedly due to the peculiar ultra-violet continuous spectrum found in many nebulae. This would probably be as good an object as any in which to study that ultra-violet spectrum.

Photographed directly this nebula appears as a disc about 15" in diameter, with a bright nucleus. The spectroscope serves to disperse the light of the nucleus and shows the nebula proper to be a hazy ring.

The intensity measurements of the images are of doubtful accuracy and are not given here. The lines present, arranged in order of intensity, are: N,, N2, HB, Hy, H8, 3869, He, 3727. The nucleus spectrum is continuous.

A beautiful example of a stellar nebula. All of the images appear as points except 3727A, which has a suggestion of haziness about it. There is no record of a continuous spectrum. The intensity measurements are not accordant, and are not recorded. The system of measurement is hardly applicable to images as small as these. The lines arranged in order of intensity are: N1, N., 3869 Hß, Hy, 3967, 4363, 4686 H8, 3727. N, and N, are very bright, relatively twice as bright perhaps with respect to Hẞ as they are in N.G.C. 7027, where they are in fact exceptionally strong in this same sense.

=

=

This is an exceedingly small nebula. The monochromatic images of correct intensity have a diameter of about 3", and as the "seeing" was very poor when the exposure was made it may

10 Sitz. Heidel. Akad. Wiss., 35 Abh., 1911.

be that the object is stellar. N, and N, are very bright relatively to Hẞ. The photometric observations are uncertain on account of the small size of the image, the great range in intensity and the position of these lines in the spectrum. Subject to these sources of error, the ratios of the intensity of N1 : N2 : Hẞ are as 380: 230: 50. The nebular lines in order of intensity are: N1, N2, Hẞ=3869, Hy, (3967 + He), H8, 4363. There is no continuous spectrum.

2

A small bright irregularly elliptical nebula. The images are all of about the same size except that at 3727A, which is somewhat larger than the others and very diffuse and hazy.

The spectrum of the nucleus is continuous, except that an emission band is suspected on the stronger plate (No. 837) at or near 4650A.

Reproductions of spectrograms numbers 723 and 727 are given in plates" XLII and XLVIII. Spectroscopic observations of this nebula are numerous. It appears to have attracted an amount of attention rather out of proportion to its interest or importance. At any rate, it has been studied to the exclusion of brigher and more interesting objects. The fact that it is comparatively large and suitable for general observation with small telescopes has probably contributed to this.

The nebula is really quite faint, and an unusually long exposure is necessary for a good record. Spectrograms have been secured with both the slitless spectrograph and the lowdispersion slit instrument. It is too large for effective observation with the present slitless spectrograph and it is hoped that a more suitable optical combination will be available for its study at some future time. It has therefore not been attempted to make the present series of plates in any sense complete.

11 The continuous spectrum in figure 4 of plate XLII comes from the moonlit sky.

The most interesting of the older observations from the point of view of the present work are those of Max Wolf12 and Burns.13 To the former we owe the discovery of the non-homogeneity of the nebula. He found the 4686 ring to be the smallest, that at 3727 to be the largest, and assigned two intermediate sizes to the others. He observed in addition that the area enclosed by the ring is relatively weak in 3727 light and strong in that of 4686A. Wolf's observations were made chiefly by the laborious process of securing successive exposures of great length with the slit of his spectrograph in different position angles, though he appears to have made some use of the slitless spectrograph. Burns' investigation was undertaken with a small slitless spectrograph attached to the Crossley reflecting telescope, and is in many respects similar to mine. His dispersion was, however, less than a fifth that of the present instrument, which is itself, as I have indicated, too low for the best observation of this object.

The image at 4686A differs from the others in the following particulars:

1. It is smaller.

2. It is almost circular.

3. It is a nearly uniform disc with those edges corresponding to the flat sides of the nebula slightly reinforced. That is, the edge of the disc is slightly brightened on the sectors corresponding to the brightest parts of the nebula.

The image at 3727A differs markedly from the others in that it shows vastly more detail. It is doubtless due in a great measure to this fact that a photograph of the nebula exhibits much more structure than is visible in the telescope. The other images are very nebulous or hazy, while this one is made up of what appears to be a complicated system of interlacing filaments, suggesting the great detail in the network nebulae in Cygnus. Wolf and Burns have found this image to be larger than the others, and while this is undoubtedly the case it is difficult to compare the sizes quantitatively on account of the unequal distribution of the light in the different images. The difference in length of the major axes, for images of normal intensity, is not very great, but the difference in length of the minor axes is considerable. The following are the measures of plate No. 727 for inner and outer diameters of three of the rings.

Measurements of the shorter diameter (along the length of the spectrum) can not be directly compared with one another in view of certain small field distortions inherent in this type of instrument which have not yet been quantitatively determined, and these diameters are therefore not given, though it is doubtless in that direction that the image 3727A has its greatest excess over the others. I am inclined to think that with the two exceptions mentioned the images are all essentially of the same size and shape. There appear to be two very faint hazy protuberances on the outside edge of the 3727 image just to the north of the northeastern end and to the south of the southwestern end of the major axis.

Dr. Burns extended his observations to the red end of the spectrum, and reports that the Ha image differs from the other hydrogen images and bears a close resemblance to the one at 3727A. As we should not expect Ha to give an image differing greatly from those of the other hydrogen lines, it seems probable that the reported Ha image is in reality due chiefly to one or both of the nebular lines 6548 and 6583, which, together, are quite as strong as Ha in some nebulae.

12 Vierteljahrschrift Astron. Gesell. 43, 283, 1908.

13 Lick Obs. Bull., 6, 92, 1910.