Remarks Many observers, including d'Arrest, Slipher, Campbell, Moore, Paddock, and Green, have been unable to see bright lines in the spectrum. In Harvard list 56, from erroneous translation of d'Arrest's observation; not in list 76. Long exposure should be made to determine whether faint hydrogen lines and the line at 3726A may not be present, as they are relatively prominent in the outlying parts of the Orion nebula d'Arrest, Moore, and Green saw no bright lines in the spectrum. Harvard list 56, from erroneous translation of d'Arrest's observation; not in list 76 In This is in NGC as a separate nebula. It is undoubtedly a part of the great nebula in Orion, surrounding star Bond No. 734, and Campbell and Moore's observations of it are described under the heading “Orion nebula,'' above. The Hẞ bright line is strong on the photographs, but N, and N2 are not visible In Harvard list 56, not in list 76. Wolf at Heidelberg reported bright In Vogel's list, Frost's Scheiner, 232. In Harvard list 56, not in list 76. Harvard, in list 76, describes this object as a cluster with stars and Moore, Paddock, and Selga observed continuous spectrum, but no bright lines Moore, Paddock, and Selga observed very faint continuous spectrum, but no bright lines Campbell and Paddock could see no spectrum. N, and N, certainly not Moore and Selga observed strong continuous spectrum, but no bright lines Harvard list 76 describes spectrum as resembling that of n Carinae. Paddock observed continuous spectrum, but no bright lines Moore observed faint continuous spectrum, confirming Harvard list 33. NGC describes it as suspected planetary. Harvard, 33, describes spectrum NGC 4649 12 38.6 +12 6 NGC 4660 12 39.5 +11 43 2 Remarks In Vogel's list, Frost's Scheiner, 232. In Harvard list 56, not in list 76. In Harvard lists 56 and 76. Wilson at Santiago could not find and confirm NGC description suggests planetary form. Curtis with slitless finds spec- Harvard list 33 describes spectrum as apparently gaseous and continuous. In Harvard list 56, not in list 76. Curtis's Crossley photograph shows it to be a spiral nebula Wolf describes the spectrum as strongly gaseous, V. J. S., 49, 161, 1914. Campbell and Paddock, observing under good conditions, saw continuous spectrum, but no bright lines In Harvard lists 56 and 76, but Harvard comment in latter list makes NGC describes it as planetary, but Curtis's direct photographs do not Campbell and Paddock could not see the spectrum of this large and well NGC describes it as "almost planetary." Curtis with slitless was unable In Vogel's list, Frost's Scheiner, 232. In Harvard list 56, not in list 76. Moore and Selga observed faint continuous spectrum, but no bright lines SPECTROGRAPHIC EVIDENCE OF ROTATIONAL AND OTHER EFFECTS IN THE PLANETARY NEBULAE RÉSUMÉ The results of our survey of forty-three planetaries, with high-dispersion spectrographs, for evidence of internal relative radial velocities, in terms of the forms of these objects, are briefly as follows: 50 The term "stellar" is usually applied to these objects, as they are indistinguishable from stars in telescopes of small aperture and low power. The image of none of these, however, is strictly stellar, as shown by an examination of the smallest objects of this class, by Messrs. Aitken and Moore, with the 36-inch refractor; but possibly an exception will need to be made in the case of N.G.C., 2621, R.A.10h 56", which Wilson could not distinguish from a star. Under good conditions of "seeing" and using magnifications from 500 to 1000, all of the northern objects were readily distinguishable from the stars, as their disks were in no case less than 2 seconds of are in diameter. In addition to the above, 3 planetaries were tested with low dispersion. Two of these show definite evidence of internal motion, and for the other it is strongly suspected. The relation existing between the forms of the planetaries and the observed internal velocities is what we should expect if these objects are in fairly rapid rotation. The planetaries which exhibit internal motion are in general elliptical in outline, and those which are the most elongated yield the maximum relative-velocity effects. A striking example of this is N.G.C. 7026, the most elongated of any planetary investigated by us; it shows the greatest inclination of the spectral lines thus far observed. In 19 of the 23 planetaries showing evidence of relative motions the observed effects along the different axes are interpretable as due to a rotation of the nebula about an axis coinciding approximately with the shorter axis of figure. These results, coupled with the relationship of form described in the preceding paragraph, appear to admit of no other interpretation than that of rotation. PROBABLE MASSES OF THE PLANETARY NEBULAE Measures of the rotational velocity of a nebula enable us, on certain assumptions, to draw some interesting conclusions concerning its mass. If, for example, we assume (1) that the axis of rotation lies in a plane perpendicular to the line of sight and coincident with the minor axis of figure of the nebula, and (2) that the maximum relative velocity observed between two points on the major axis at equal distances on either side of the nucleus is twice the orbital velocity of a particle at this distance from the center; then the mass of the nucleus of the nebula may be computed for any assumed parallax of the object.51 The masses of three planetaries computed upon these assumptions are here given. All of these objects seem to be much more massive than our solar system. Moreover, it must be remembered that the results obtained in this manner are minimum values of the masses, since in general the axes of rotation will not be at right angles to the line of sight. One other factor, and possibly two or more factors, may also be operating to reduce the masses thus computed v2r k2 51 This is readily computed from the familiar formula for a circular orbit, M + m =- where M and m are respectively the masses of the nucleus and the particle in terms of the Sun's mass, r the radius of the orbit in astronomical units, v the linear velocity of the particle in astronomical units per day, and k the Gaussian constant. 52 Lick Obs. Bull., 9, 100, 1917. 53 Publ. A. S. P., 29, 209, 1917. 54 Proc. N. A. S., 3, 133, 1917. below their real values. If there is radiation pressure between the matter in the nucleus and the outlying molecules whose differential velocities have been observed, then those velocities are lower, and the computed mass of the nucleus will be correspondingly smaller, than if gravitation alone were concerned in maintaining equilibrium. What shall we say of the effect of an expansive gas filling the space between the nucleus and the outlying molecules? Does such a gas exist? If so, is its action on the outlying gas molecules analogous to what we should expect from a strong expansive coiled spring operating to separate farther the two components of a double star, in effect to depress their resultant mutual attraction below that due to their real masses? We have considered the probable masses of only three nebulae, above. In analogous manner we could discuss the probable order of mass magnitudes possessed by the other 16 planetaries whose rotational effects we think we have observed. If we assume reasonable distances for them, their masses all come out several times the Sun's mass. If we assume that each mass equals the Sun's mass, the distances all come out unreasonably small. We conclude, therefore, that the minimum probable masses of the planetary nebulae in which we have observed rotational effects are several fold the Sun's mass. If these particular nebulae ultimately develop into solar systems, we should expect those systems to be much more massive than our solar system. However, we have been successful in measuring rotational effects in only those planetaries, in general, which have considerable angular diameters and the greater intrinsic brilliancies. It is probable that the apparently smaller and fainter planetaries are on the average really the smaller and less massive members of their class; and we have no valid reason for doubting that many planetaries may in mass be comparable with, or smaller than, our solar system. That the planetaries are not rotating as solid bodies is indicated by the S-shaped lines which we obtain in their spectra; demonstrating, we believe, that different strata have different speeds of rotation. Evidently, the interpretation of our results for relative velocity are complicated by this fact. The planetaries are undoubtedly transparent to considerable depths, and we are receiving light, at any point in a spectral line, from strata of different rotational speeds. OTHER EFFECTS THAN THOSE OF SIMPLE ROTATION For four nebulae, of which N.G.C. 2392 is the extreme case, the forms and inclinations of the lines are not such as to be readily interpreted as simple rotation effects. We refer especially to the doubling of the lines from the central areas of these nebulae, the components of the lines. showing considerable variations in relative intensity in different parts of the nebulae. In addition to the 4 objects in which this effect appears so prominently, 6 other planetaries show this phenomenon in undoubted combination with rotational effects. Of this latter group, N.G.C. 7662 is typical. Following is a list of the 10 planetaries in which we have observed the doubling of the nebulium lines, with the spectral sub-class assigned to each in the Harvard classification of the spectra of gaseous nebulae (Class P). Eight of these planetaries belong to Class Pe, those in which the line 4686A, thought to be due to helium, is present. Nine objects of Class Pe have been observed with high dispersion. 55 Annals H. C. O., 76, 19, 1914-1916. The one for which we have not obtained the effect (N.G.C. 2440) is faint and not well situated for the required exposure. Our tests for this one are accordingly insufficient. The objects which show best the doubling effect in the spectrum lines are in general those with very bright ring formations or strong condensations. We have not found this effect in planetaries whose images appear as disks of nearly uniform intensity. A typical example of the latter condition is N.G.C. 6826, for which the spectrum lines show no trace of broadening or division into components. The lines seem to be monochromatic and to have the same inclination through practically their whole length. The bright condensations in Curtis's drawing50 of this nebula have been shown by Mr. Wright to be emitting the ultra-violet radiations, and they are not visible in the N,, N., and many other images. Our results seem to indicate that this phenomenon of double lines is dependent upon the same physical conditions in the nebulae that are favorable for the appearance of the line 4686A and for the existence of well defined bright rings in the nebular structure. A number of hypotheses to account for the doubling and peculiar details of the lines in this group of planetaries have been examined by us: motions of expansion or contraction, i.e., along radii of the nebulae, in combination with motions of rotations about an axis; the Zeeman effect; the Stark effect; pressure in the deeper nebular gases, combined with absorption by the outer strata; anomalous dispersion; the various combinations of these sets of assumed conditions. We have found no hypothesis which seems competent to explain all of the observed facts. It is evident that an interpretation of this phenomenon from the standpoint of radial velocities alone is attended with many difficulties: we are unable to imagine a simple ring system, or an arrangement of vortex rings, in harmony with the forms of these lines. A fairly transparent contracting ellipsoidal shell, in rotation, might give rise to double lines, but not with the relative intensities observed in the components of the nebular lines. The series of experiments made on N.G.C. 7662 in an attempt to detect plane, circular, or elliptical polarization in the components of the double lines, furnish strong evidence that the components are not polarized, and that this phenomenon is not a simple Zeeman effect. As to the Stark effect in these nebulae, our experiments are not adequate to prove its presence or to disprove its existence. Laboratory experiments seem to indicate that the Stark effect is widely different for different spectral lines, and the tests for it are not so definite as for the Zeeman effect. There is some indication that the observed effect may arise from the broadening of the central section of the line, due either to particles moving with different radial velocities or to slightly higher pressure of the gases in the central region of the nebula, and that the dark medial line is simply a reversal produced by the absorption of an outer stratum of material which has a reduced rate of rotation. If the central section of the line is inclined, due to rapid rotation of inner nebular structure, the absorption line, on this hypothesis, would show less inclination, and such is the observed fact. We should also expect to find the violet component stronger than the red one on one side of the nucleus and weaker on the opposite side. That is precisely what we do observe. We might also expect to find the red component of greater intensity than the violet, since the effect of pressure is in general both a broadening of the lines and a shift of the maximum of intensity toward the red. This effect would be particularly noticeable in the lines obtained along the minor axis of the nebula, and we do observe the red component to be the stronger through the greater part of the central section. Although this hypothesis represents the observed facts in a general way, it must be admitted that it meets with some difficulties when we attempt to explain all of the observed details. Part of the peculiar appearance of the lines may arise from the fact that we are receiving light from very different depths, which probably have different rotational speeds. If we are 56 This volume, pl. XX, fig. 58. |