The interference method has the great advantage of yielding radial-velocity results simultaneously for as many points as the observer pleases to select within a given area-in this case a circular area 4' in diameter-whereas the ordinary spectrograph gives velocities for the exceedingly restricted area which coincides with the narrow slit. It seemed to us important that the interesting results obtained by the new method should be confirmed by observations made in the usual spectrographic way, and especially that the fainter parts of the nebula, at considerable distances out from the central region, should be observed for radial velocity. As we were then engaged upon our extensive programme of measuring the radial velocities of all available brightline nebulae, we accordingly began upon the Orion nebula early in November, 1914, as soon as that object came into good observing position. We have continued the observations on the Orion nebula, as opportunity offered, through four successive winters and have recently completed them. The observations are described at the proper point in this paper. The desirability of confirming the interferometer results appealed also to Director Frost of the Yerkes Observatory, and accordingly Messrs. Frost and Maney secured seven 1-prism spectrograms of the brighter regions of the Orion nebula, in January-March, 1915. Their observed velocities vary from +6.0 to +19.6 km. per sec., and agree with ours in a general confirmation of the Marseilles results. Professor Keeler, in 1891, looked for a rotation effect in the spectrum of N.G.C. 3242 (G.C. 2102, R.A.=10h 20m), with negative results. In his opinion a difference of velocity for two parts of the nebula could not have been detected with his apparatus if this difference were less than 12 or 13 km. per sec. (p. 203). In N.G.C. 6572 (G.C. 4390, R.A.18h 7m) the principal lines were observed by Keeler to be fine and sharp in the spectrum of the nebulous envelope, but fuzzy and considerably broadened in the nucleus (p. 209). A similar observation was made by Keeler for N.G.C. 6210 (G.C. 4234, or 25, R.A. = 16h 40m) (p. 220). = Professor Hartmann, in 1901, sought for spectrographic evidences of relative motions within the planetary nebulae N.G.C. 6542 (G.C. 4373, R.A. 17h 59m) and N.G.C. 6572. His results suggest that the north edge of N.G.C. 6543 is approaching us 9.4 km. per sec. more rapidly than the center of the nebula, and that the west edge of N.G.C. 6572 is approaching us 3.8 km. per sec. less rapidly than the center; but Hartmann is not inclined to affirm that the apparent differences are real, as some of the plates were secured with a dispersion of only one prism and most of the plates were under-exposed. Our own observations do not confirm the existence of the differences described. Shortly following the realization that our spectrograms secured with the 3-prism and 16-inch camera combination were yielding radial velocities of great accuracy, and that nebular lines corresponding in length to the full diameters of the brighter planetaries could be recorded with this high dispersion, using exposures of reasonable length, we determined to test several nebula for evidences of rotation. The planetaries are in general elliptical in outline, and it was natural to assume that any rotational effect would be observed as a maximum along the major axis of figure. The slit should accordingly be set upon the major axis of the image, and the central star, or nucleus, of the nebula should be held upon the middle point of the slit during the exposure. If the recorded spectrum lines of the nebula are found to be inclined to the zero direction, as supplied by the lines in the comparison spectrum, it is necessary to consider that this inclination. may be due to: (1) A rotation of the nebula; (2) A non-uniform illumination of the slit; (3) Faulty instrumental equipment or adjustments; or (4) Two or three of these factors in combination. 7 Pop. Astron., 53, 485-7, 1915. 8 Astroph. Jour., 15, 287, 1902. A little consideration will show that a comparison of two spectrograms obtained with the image on the slit respectively direct and reversed 180° is in general sufficient to differentiate or to eliminate the effects (2) and (3), if any such exist. If the cause (1) is operating, the inclination of the nebular lines to the zero direction should be reversed for the two positions of the slit. If the ellipticity of the nebular image is controlled wholly or in large measure by the dynamics of rotation, we should expect that a spectrogram secured with the slit in coincidence with the minor axis would show no inclination of the lines. The first nebula studied in this manner was N.G.C. 7009 (R.A. = 20h 50m), in November, 1915, using dispersive power supplied by three dense 60° prisms (the prisms of the original Mills spectrograph) and a 16-inch camera, and a slit width of 0.075 mm. The nebular lines were found to be markedly inclined to the zero direction. Two or three spectrograms obtained with the slit coincident with the major axis of the nebula, but with the image in positions on the slit respectively direct and reversed 180°, confirmed the results of the first observation. fourth spectrogram, taken with the slit coincident with the minor axis of the nebula, gave no certain evidence of inclination of the lines. These and other similar observations will be described in their proper order below. THE OBSERVATIONS The following pages describe the observations made at Mount Hamilton and at Santiago, in the order of the right ascensions of the objects. The reader will have the benefit of the textual descriptions and the admirable photographs or drawings of the northern nebulae made by Dr. Curtis, in a preceding part of this volume, and we do not attempt to supply many additional items of description. The descriptions of the nebulae south of Decl. -34° are by Dr. Wilson. The right ascensions and declinations and the equivalent celestial latitudes and longitudes of the nebulae, as quoted, are for the epoch 1900.0. The first column contains the Greenwich mean time of mid-exposure. The second column contains the duration of exposure expressed in hours. The combination of prisms and cameras employed is described in the third column. The collimator section of the original Mills spectrograph has been used in all Mount Hamilton observations. The plates were measured by the observers listed in the fourth column. The bright lines measured are indicated in the fifth column. The radial velocities obtained by individual observers are quoted in the sixth column. The relative weights assigned to the results from the individual spectrograms of each nebula are in the eighth column. The length of the slit and the position angle of the slit as projected upon the sky are described in the ninth column. In this column the word "long" signifies that the slit was made long enough to include the entire diameter of the image of the planetary nebula concerned, or, in the case of the Orion nebula, that the slit was about 100" long. It should be explained that when a "stellar" planetary nebula was under observation, a "long" slit was in absolute measure quite short. With slits of "medium" and "short" lengths no attempt was made to have the slit extend entirely across the image of the nebula under observation, or to "follow" with great accuracy. We have used a variety of slit widths, depending upon the instrument employed, the brightness of the object, and the special purpose for which the spectrogram was obtained. It seems not worth while to quote the slit width for each spectrogram, but we describe our general practice at this point. In the early observations made with the 1-lt. pr. 16-in. combination, the slit width was usually 0.125 mm. With the 1-pr. 16-in. combination, the slit width was usually 0.10 mm., and occasionally 0.088 mm. With the 3-pr. 5-in. combination, the slit width was nearly always set at 0.25 mm., but on a few fainter objects the width was 0.375 mm. With the 3-pr. 6-in. combination, the slit width was usually 0.20 mm., but on a few of the fainter objects it was 0.30 mm. With the 3-pr. 16-in. combination, we nearly always used a slit width of 0.075 mm., but on a few objects, especially the fainter ones, it was 0.10 mm., and on some of the brightest nebulae it was as narrow as 0.05 mm. With the 3-pr. 32-in. combination, the slit width was usually 0.075 mm., but on some objects it was reduced to 0.05 mm., and in a few cases, in order to obtain very narrow nebular lines, the width was 0.038 mm. We have neglected to secure accurate data from Chile as to the precise slit widths employed, but they did not differ greatly from the slit widths at Mount Hamilton. The last column of observational data, "Remarks," contains statements for the most part self-explanatory. Many comments on the degree of exposure are included, in order to assist future observers in arriving at an approximate estimate of the brightness of the nebular lines in the various objects, and to serve as a guide to the system of assigning relative weights. Such ratios as 10:3:2 in the "Remarks" column are the estimated relative photographic intensities of the bright lines listed in the "Lines" column. When the slit and correspondingly the nebular lines were long, micrometer readings were obtained for several sections of the long lines; for example, on the last two spectrograms of N.G.C. 40, R.A.=0h 76. For the planetary nebulae the radial velocities obtained from the several sections have usually been combined into one mean observed velocity, as in the case quoted; but for the Orion nebula the radial velocities supplied by individual sections are sometimes quoted, but again, and for the most part, the results from two sections or three sections are combined into one mean result for printing and further use. The radial-velocity observations of bright-line nebulae made by Keeler, and the very few made by others, are quoted in their logical places; but we have not modified their results to conform to the system of wave-lengths used by us. The list includes 138 nebulae with bright lines in their spectra : (a) 125 planetary and irregular nebulae whose radial velocities we have determined at Mount Hamilton and at Santiago; (b) 9 planetary and irregular nebulae whose spectra contain the characteristic bright lines, but which are so faint that we have not found it practicable to make the very long exposures that would be required to record the lines in measurable strength; (c) 4 spiral nebulae whose spectra contain the characteristic nebular bright bands in addition to the usual continuous" spectrum of spirals. 66 No doubt the future will extend the list appreciably, especially by the addition of objects in the southern sky. All of the known planetaries in the northern heavens which are of sufficient brightness for observation with high dispersion have been tested for evidence of internal relative velocity, according to the method outlined above. The results of this investigation for any particular object are described in the paragraphs immediately following the tabular data referring to the general radial velocity observations of that object. In order to exhibit the complicated forms and interesting detail shown by the spectral lines of certain nebulae, drawings of the N, line were made by Mr. Moore from the best spectrogram of the particular nebula. For some of the objects a direct enlargement of the nebular line is also reproduced, together with the drawing of the line, as a check upon the relative intensities and general form exhibited in the latter. In the drawings, which are all made to scale from actual measures of the lines, the prismatic curvature has been removed. A weale placed at one side of the line represents the distance from NGC 10 a 0 76; 8-471° 58′; λ=51° 25'; B+60° 15' The spectrum of this object has several points of special interest. The radiations from the nebulosity consist almost exclusively of the Hẞ and Hy hydrogen lines, in the region photographed. The N, line is very faintly recorded on the last plate, exposure 31.4 hours, and there is a scarcely recognizable though certain trace of N, on the same plate, but N, and N, are not visible on any of the other plates. All measures for radial velocity are based upon Hẞ alone. The central star in N.G.C. 40 is one of the brighter nuclear stars in the known planetaries. Professor Pickering announced in 1905, in Harvard Observatory Circular, No. 98, that the spectrum of this nebula resembles the Wolf-Rayet type; and in 1914 Dr. Paddock called our attention to the fact that the nuclear star is of the Wolf-Rayet type, as indicated by his measured wave lengths of the broad bright bands in the 4600-4700A region. The HB line in the nebula is not a "straight" line. The last plate, with the slit upon and a little longer than the major axis of the nebula, records Hẞ to the full length of this axis. The helium comparison lines are of course concave to the violet, as the natural result of prismatic dispersion, but the section of the nebular Hẞ line corresponding to the diameter of the nearly circular central structure, 38" long, is slightly convex to the violet on its violet edge and almost rectilinear on its red edge (see pl. XXVII, fig. 4, 1917, Sept. 22). The section of Hẞ corresponding to the north ansa of the nebula is offset perceptibly to the violet. However, the character of the line must not be described too minutely, as it is under-exposed for the ansae and the regions immediately interior to them. In measuring some of the plates, much depends upon the part of the line selected for bisection. There is a slight indication that the SSW edge of the brighter structure, 35" by 38", is approaching us a little more rapidly than the NNE edge, thus indicating a slow rotation of the nebula, the N and S ansae not showing this effect; but the dispersion is really too weak and the under-exposure of the extremities of the line too severe to justify a positive statement on the subject. * Campbell measured only the outer ends of the line Miss Hobe measured the central section. * Compbell mexxured the end sections of the line - Miss Hobe's measures refer to the greatly strengthened central point of intersection with the strong Continuon spectrum of the nuclear stat. |