O VI AND He II EMISSION LINES IN THE SPECTRA OF QUASARS

Varshni,Y.P: 1977, Astrophys.Space Sci. 46, 443.

Also available in Adobe Acrobat PDF format.

Hydrogen Alpha Images of Planetary Nebula
IC 1747	Abell 30	NGC 246
NGC 2371	NGC 2452	NGC 7026
NGC 6905	NGC 5189	NGC 2867

Abell 78 courtesy of M. Horn, the faint outer halo is composed of normal stellar material (mostly hydrogen) and the bright inner elliptical ring is composed mostly of helium.

1. Introduction

1. There is no red shift, and
2. The strength of the emission lines is due to laser action (more specifically, due to amplified spontaneous emission).

The proposed model resolves many of the enigmas commonly associated with quasars, and is consistent with various pieces of evidence (Varshni, 1974b, c, 1975a, b, 1976a, b). Varshni and Lam (1976) have presented model calculations for laser action in He II λ4686 within the framework of the PLS model.

In the present paper we further develop the thesis that the quasars are stars. We demonstrate the continuity of the spectra of ten quasars with those of O VI sequence planetary nuclei and stars. Although the present paper deals with this particular spectral class of quasars, certain general problems concerning the spectral classification of quasars are also discussed. We identify O VI λλ3811, 3834 and He II λ4686 emission lines in the spectra of these ten quasars. Candidate identifications for other quasar lines are also suggested. Making use of the similarity between the spectra of O VI sequence planetary nuclei and those of the ten quasars, we estimate the absolute magnitudes, temperatures and masses of these quasars. The distribution of quasars in galactic coordinates is also discussed. Three predictions are made.

2. O VI Sequence Planetary Nuclei and Stars

TABLE I.   O VI emission lines in Wolf-Rayet spectra
Author(s)	Star		Intensity (arbitrary units)	Proposed identification
Wright (1918)		3815
Plaskett (1924)	HD 192103	3813.5	3
Beals (1930)		3815	3
		3835	2
Aller (1943)	HD 192103	3815	3.0	He I λλ3819.6, 3819.7, O VI
		3835	2.9	O VI, He II λ 3833.8
	HD 193793	3815	3	He I λ 3820, O VI
		3835	absent(?)
Smith (1955)	HD 119078	3809.7		O VI
	(WC7)		comparable
		3836.4		O VI
	HD 115473	3811	21.8	O VI
	(WC6)
		3834	21.6	O VI
Underhill (1959)	HD 192103	3814.6	medium	He I, He II, O VI
		3833.8	medium	He II, C II, O VI, S III
Aller and	γ2 Velorum	3811	3.0	O VI
Faulkner (1964)		3835	2.8	O VI

We list these 12 planetary nuclei in Table II (category A) in order of increasing strength and width of O VI lines (as arranged by Smith and Aller, 1969). The widths of the emission lines range from about 40 Å (e.g., NGC 7026 and NGC 5189) to about 1 Å (NGC 246); when the lines are broad, the two components of the O VI doublet are blended together, as are the He II λ4686 and C IV λ4658 lines. In Table II for each nuclei we also give its equatorial and galactic coordinates, distance (d), apparent magnitude (m), absolute magnitude (M), and absolute magnitude taking into account space absorption (M(a)). Distances are from Cahn and Kaler (1971); where several values were quoted for the same nuclei, an average was taken. Apparent magnitudes are from Smith and Aller (1969). The M(a) values were calculated from the formula

In order to obtain A it was assumed that the absorption of star light is 1.9 mag/kpc (Allen, 1973). The scale height for absorption above galactic plane was taken to be 140 pc (Allen, 1973) and the same value was assumed for the scale height of absorption below the galactic plane.

It was pointed out by Sanduleak (1971) that this unusual variety of Wolf-Rayetlike spectrum (O VI sequence) also occurs in stars that do not appear to be planetary nuclei. He provides a list of five such stars. Here we shall label these stars as Sand 1, 2, 3, 4 and 5. Coordinates and magnitudes of these stars are given in Table II (category B).

In Figure 1 we show a diagrammatic representation of the spectra of three of the planetary nuclei namely NGC 6905, NGC 7026, and NGC 5189, and two of the stars, Sand 1 and Sand 4, belonging to the O VI sequence. Only the strong lines have been shown. The sources of data and some additional information about these objects and also Sand 2 are given below.

• NGC 6905 and NGC 7026. Emission-line spectra of nuclei of both of these nebulae have been obtained by Smith and Aller (1969). Aller (1968) provides the tracings for the range 3200-4700. Recently, Aller (1976) has observed the central stars of NGC 6905 and NGC 7026 with the Lick Observatory image-tube scanner. He gives scans of NGC 6905 for the ranges 5270-6500 and 6650-8500, and those of NGC 7026 for 5270-8400.
• NGC 5189. The spectrum of the central star in NGC 5189 has been reported by Blanco et al. (1968a) and by Smith and Aller (1969). Both studies cover the spectral range shortward of 5050 Å. Possible variability in the relative strengths of the two features has been discussed by Sanduleak (1968).
• Sand 1. This star was discovered by Miss Cannon (1933) in the vicinity of the Small Magellanic Cloud, and classified as being of O type and having a broad emission feature which apparently she interpreted as He II 4686. Westerlund (1964) reclassified this star as W-R+O. Sanduleak (1968) obtained a spectrum of this star and found two broad emission bands centred at 4670±5 and 3840±5 Å, respectively. He gives a microphotometer tracing of the spectrum for the range 3350-4900.
• Sand 2. Discovered by Sanduleak (1969, 1971) and designated as star -68°145 in the catalog of Large Magellanic Cloud members.
• Sand 4. Discovered by Blanco et al. (1968b). Its spectrum was first discussed by Blanco et al. (1968c) and later by Freeman et al. (1968), who give an excellent microphotometer record of the spectrum for the range 3700-7300. Possible variability in the strength of the emission lines has been reported by Johnson and Golson (1968). In some of these references, the star has been referred to as GX3 + 1. Cohen and Barlow (1974) have observed a lower limit of 4.1 for the 10 micron (infrared) magnitude of this star. There is an associated nebula with this star, which, according to Johnson (1975), 'should qualify as a supernova remnant except for the complication that it contains a peculiar, hot star'. The radiofrequency spectrum from the nebula appears to be nonthermal (Johnson, 1973).

Fig.1. Diagrammatic representation of the spectra of the following: (A) Central stars of three planetaries, namely, NGC 6905, NGC 7026, NGC 5189, (B) Stars Sand 1 and Sand 4, and (C) Ten quasars. The code for the strengths of the lines is explained at the top of the diagram. The wavy vertical lines indicate, where known, the limits(s) of the observed spectrum. For the quasar 4C 29.68, the figure shows the set (c) in Table III; for 3C 48, it is a composite figure from the two sets of data, and for 4C 37.43, the set (a) has been shown.

The spectrum of the central star of NGC 246 (Heap, 1975) affords an excellent illustration of the O VI lines; the lines are relatively sharp and their relative intensities appear to be close to the theoretical values (see tracing in Heap, 1975). We note that the nebula is so faint that there is practically no interference by nebular lines on the stellar spectrum.

However, in some other cases the situation is not so clearcut. It is observed that in some cases where the lines are broad, 3834 is of comparable intensity to 3811 (e.g., NGC 6905, NGC 7026, see tracings in Aller 1968, also Table I of the present paper). When the lines are so broad that they are completely blended, it seems that 3834 becomes stronger than 3811. This statement is based on the following evidence. Sanduleak (1968) has measured the position of the centre to be 3840±5 Å in Sand 1, and the microphotometer tracing for the spectrum of Sand 4 given by Freeman et al. (1968) shows the peak to lie at 3850 Å.

When the O VI lines are strong and broad it would be reasonable to assume that the underlying mechanism is the same as that we have proposed for emission lines in quasar spectra (Varshni, 1974a, 1975a; Varshni and Lam, 1976). To explain the increasing strength of 3834 with the width, the following possibilities are open.

1. Under non-LTE conditions the populations of P_3/2 and P_1/2 levels need not be proportional to their respective statistical weight. It is possible that the population of the P_1/2 level becomes comparable to, or greater than that of the P_3/2 level as departure from LTE conditions increases.
2. There is another line or multiplet at 3835 which undergoes laser action when there is high speed expansion. A possible candidate for this alternative is multiplet 5 of S III (3832-3838).

3.1. The Data

The spectral data for these ten quasars are summarized in Table III. For each quasar the first row shows the reported wavelengths. Below each wavelength, where known, are indicated the strength and width of the line. The following abbreviations have been used. Strengths: VS=Very Strong, S=Strong, M=Medium, Wk=Weak. Widths: wd=wide (40-100 Å), mwd=medium (20-30 Å), n=narrow (10-15 Å). The quantitative equivalents, shown in parentheses, are from Burbidge and Burbidge (1967).

TABLE III.  Emission line data for quasars. Below each wavelength are given, where available, the strength and the width of the line.
Quasar (red shift)	Set	1	2	3	4	5	6	7	8	9	10	11
DW 0202+31		3823		4708
(1.466)		S, mwd		M, wd
PKS 2145+06		3824		4682
(0.367?)		Wk, wd		Wk, wd
PKS 2059+034		3840		4686
(0.370)		M, wd		Wk, wd
4C 33.03		3808		4687								6861
(1.455)		Wk, wd		Wk, wd								Wk, wd
PKS 0906+01		3843		4695			5662
(1.018)		Wk, wd		Wk, wd			M, wd
4C 29.68	(a)	3846					5639
(1.015)		M, 35 Å					Wk, 35 Å
	(b)	3834					5625
(1.009)
	(c)	3841		4675			5645
(1.013)		Wk, 95 Å		Wk, mwd			Wk, wd
4C 27.38		3837	4591	4700	5114	5310	5631
(0.372)		S, 10 Å	Wk, 10 Å	Wk, 10 Å	M, 10 Å	M, 10 Å	M, 10 Å
4C 09.31		3822		4681	5088	5285				6628	6774	6841
(0.366)		20 Å
4C 48	(a)	3832.3	4575	4685	5097	5288		5935	6349	6646
(0.3675)		S, 70 Å	Wk, mwd	M, mwd	M, n	Wk, n		S, mwd	Wk, mwd	VS, wd
	(b)				5099.4						6781.7	6849.6
(0.3679)
4C 37.43	(a)	3837		4691		5301		5969		6640	6799	6862
(0.370)		M, mwd		Wk, mwd		M, mwd		S, wd		M, 160 Å	M, n	M, n
	(b)	3833		4704		5301				6638	6800	6862
(0.371)		M, 50 Å		Wk		Wk				M, 100 Å	10 Å	M, 10 Å

• DW 0202+31. The data are due to Burbidge (1970).
• PKS 2145+06. We quote verbatim from Kinman and Burbidge (1967) and Burbidge (1968). Kinman and Burbidge (1967): 'Lines at 3824 and 4682 Å were measured in 2145+06, and these are probably Mg II 2798 and [Ne V] 3426, with a red shift z = 0.367. However, H beta should then appear at 6643 Å, and we have not seen this line on spectrograms taken in the red with the image-tube spectrograph. This possible red shift needs confirming.' Burbidge (1968): 'Kinman and Burbidge (1967) obtained several spectrograms of this object and measured broad emission features at 3824 and 4682 Å. We tentatively identified these as Mg II 2798 and [Ne V] 3426 at z = 0.367, but H beta, which should have appeared at 6643 Å, was absent. Therefore, we did not include this object in our table of red shifts. Two further spectrograms have been obtained, one with each spectrograph. The two lines are less distinct, although 3824 is certainly still visible, and again no other features can be seen. The red shift is thus still not definitely determined.'
• PKS 2059+034. The spectral data are due to Burbidge and Strittmatter (1972). Photographic spectrograph was used which covers the spectral range 3200-5000 Å.
• 4C 33.03 (3C 48/54). The spectrum of this quasar has been reported by Burbidge (1970).
• PKS 0906+01 (4C 1.24). Data are due to Burbidge and Strittmatter (1972). Both the photographic spectrograph (spectral range 3200-5000 Å) and the Carnegie image-tube spectrograph (spectral range 4000-7200 Å, low resolution) were used.
• 4C 29.68 (CTD 141). Three sets of data are available.
  • (a) Schmidt (1966) observed it in the spectral region 3200-6200 Å.
  • (b) Lynds' (1967) spectroscopic material was obtained with the Cassegrain spectrograph of the Kitt Peak 84-inch telescope.
  • (c) The third set is due to Kinman and Burbidge (1967), who note that the line seen by them at 4675 Å was not measured by either Schmidt or Lynds.
• 4C 27.38. The data are due to Schmidt (1974). This quasar is one of 39 investigated by Schmidt (1974), and he states 'The spectra usually cover the range 3200-6700 Å for each object'. He also states that for this quasar 'All the lines are unresolved, i.e., less than 10 Å wide'. If this finding is confirmed by independent observations, the spectrum of this quasar could be used for accurate determinations of the emission line wavelengths, which would be of considerable value in their identification.
• 4C 09.31. The spectrum of this quasar was obtained by Lynds and Wills (1972) at the Kitt Peak National Observatory. The wavelength region covered is not given, nor are the intensities and widths of lines, except in one case where the width is noted. The wavelength values given in Table III are the calculated ones obtained from their supposed red-shifted identifications and the apparent red shift.
• 3C 48. Preliminary observations of the spectrum of this quasar were reported by Greenstein and Munch (1961) and by Matthews and Sandage (1963). Greenstein and Matthews (1963) interpreted the spectral lines on the red-shift hypothesis and a detailed study was made by Greenstein and Schmidt (1964). The first set of data given in Table III are from the last named source. Greenstein and Schmidt (1964) have given numerical estimates (on a relative scale) for the intensities of emission lines. For comparative purposes we have converted them into qualitative measures like 'strong', 'medium' etc. In addition to the lines recorded in Table III, three very weak lines at 4065.7, 4205.3 and 5136 Å, respectively, are also present. Ford and Rubin (1965) obtained the spectrum of 3C 48 with the aid of a cascaded- image tube using the Perkins 69-inch telescope at Lowell Observatory. They give a tracing of the spectrum for the range 4200-7500 Å, however, no measured wavelengths are given. Wampler et al. (1975) have carried out observations of the nucleus of 3C 48 and also that of the nebulosity surrounding it, using the image-tube scanner (Robinson and Wampler, 1972) at the Cassegrain focus of the Lick 120-inch telescope. Their observations cover the spectral range 4800-7000. In addition to the three lines listed in the row labelled (b) of Table III, there are features in their scan of the nucleus corresponding to line nos. 5, 7 and 9; however, their wavelengths have not been given by Wampler et al. (1975).
• 4C 37.43. Two sets of data are available: (a) Burbidge (1968) obtained the spectrogram(s) at the prime focus of the 120-inch telescope at Lick Observatory. (b) Schmidt (1974), who has reported on 39 quasars in his paper, states 'The spectra usually cover the range 3200-6700 Å for each object'. It is interesting to note that a strong line at 5969 Å reported by Burbidge (1968) was not observed by Schmidt (1974), though obviously the latter's observations (Table III) cover this region. We note that this line is quite close to the night sky line 5893 and probably this fact has some bearing on the difference between Burbidge (1968) and Schmidt (1974).

3.2. Accuracy of the Data

Lynds (1967) has used an English Electric Valve P829 D image intensifier in conjunction with the Cassegrain spectrograph of the Kitt Peak 84-inch telescope. The good points and limitations of the image tube used by Lynds and coworkers are summarized by Livingston et al. (1966). For one of the quasars, 0122-00, Lynds' results can be compared with the more recent ones obtained by Browne et al. (1975) using an image-tube scanner (Robinson and Wampler, 1972) at the Cassegrain focus of the Anglo-Australian 3.9-m telescope. Lynds (1967) gives the wavelengths as 3210, 3947 (strong), and 5797 Å, respectively. The observations of Browne et al. covered the region 3573-6133 Å, and they give the wavelengths as 3968 (Line/Continuum ratio = 0.4) and 5848 Å (Line/Continuum ratio = 0.6). They also state that 5848 appears to be asymmetric, suggesting self-absorption in the blue wing. It will be noticed that there are large discrepancies between the two sets of wavelengths. Also, from the data given by Browne et al. it would appear that 5848 is stronger than 3968, while Lynds' data would appear to imply the reverse. Did the strength of one or both of the lines change during the interval between Lynds' and Browne et al.'s observations? What is the accuracy of Lynds' observations? Obviously some light could have been thrown on these questions if microphotometer tracings were published. Unfortunately, microphotometer tracings of the emission spectra of only a few quasars (out of about 400 for which data are now available) have been published.

The quasar spectra share with the Wolf-Rayet spectra the common property of having wide emission lines. We consider it appropriate to quote Edlen's (1956) opinion on this point: 'It may be remarked, by the way, that it is practically impossible to describe a Wolf-Rayet spectrum properly by means of a wavelengths list alone. An exhaustive description of a spectrum of this kind would require, in addition to the table, a spectrum reproduction and a photometer tracing of the same magnification, which should be large enough for reading off details on an adjoined wavelength scale. The reproduction of HD 192103 published by Swings (1942) sets an excellent standard with regard to magnification and gradation.' Edlen's comments about Wolf-Rayet spectra are applicable, if anything, to a greater degree, for quasar spectra.

As another example of the difficulties in assessing the accuracy and reliability of the published data, we quote the case of the quasar 2136+14 (=OX 161). Two investigations on this quasar were published in the same month, one by Wills and Wills (1974) and the other by Strittmatter et al. (1974). Both groups used image-tube spectrographs. Their data are given in the following, labelled I and II, respectively,

• I 4167 Å (strong), 4794 Å, 5303 Å
• II 4165 Å (Very strong, wide), 4260 Å (Medium, wide), 4820 Å (Weak, wide, possibly confused with night sky), 5302 Å (Medium, medium wide).

Comparisons such as the ones given above, for quasars for which two or more sets of data are available, show that:

• (a) Wavelengths of lines described as very strong or strong agree to +3 Å, unless the line is very wide, in which case the difference can be larger.
• (b) Wavelengths of lines described as medium often differ by +10 Å, differences as large as +20 Å are not uncommon.
• (c) Weak lines reported by one observer are often not reported by another observer at all. Differences in the reported wavelengths are frequently very large. For very weak lines the situation is even worse.

Night sky lines and lines from city mercury-vapour lights also create problems. If a quasar line falls near any of these lines, it is obvious that there can be considerable uncertainties in the wavelength and strength of the quasar line, especially if its strength is only medium or weak (see also Roeder and Dyer, 1972).

Allowing for the uncertainties present in the reported data that we have discussed above, the wavelengths in each vertical column in Table III may be considered to be more or less the same.

3.3. Spectral Classification of Quasars

In the present instance we require that two emission lines at about 3820 and 4686 Å respectively be present in a quasar if it is to be a member of the O VI. It will be noticed from Table III and Figure I that the spectra of the ten quasars therein satisfy this requirement. A scrutiny of Table III indicates that in some cases the peak of the O VI emission lies at about 3840 Å - we have earlier drawn attention to a similar situation in the spectra of the stars Sand 1 and Sand 4. It will also be noticed that these quasars can be divided into three groups according to their red shifts. These occur at z = 0.37, 1.015, and 1.46. We have noted in a previous paper (Varshni, 1976b) that some quasars with very different red shifts belong to the same spectral class; the present example illustrates that point. We may also note here that probably there are a few other quasars, which belong to the O VI sequence; however, the available data for these quasars are either incomplete or not accurate enough, or both.

4. Line Identifications and Comparison of Spectra

4.1. General Remarks

1. Accuracy of the observed wavelength. When a line is broad, one naturally has to allow a certain amount of uncertainty in its wavelength.
2. Extent of the available laboratory data at that time.
3. A reasonable guess as to the composition of the emission region in the astronomical source and the conditions of excitation in it.
4. Self consistency in the proposed identifications for various lines.

When the emission lines from an astronomical source are broad, their identification poses especially difficult problems. The centre of the line is difficult to determine, and if the profile is asymmetrical this difficulty is further compounded. In addition the lines are sometimes blended, further confusing the picture.

The history of the identification of 4650 in Wolf-Rayet spectra may be given here as an illustration. Plaskett (1924) thought this line was due to C II, Beals (1930) attributed it to C III. Edlen (1956) considers it to be a blend of C III and C IV. Smith and Aller (1969) identify it with C IV λ4658.

4.2. Line Identifications and Discussion

• Line no. 1. O VI doublet.
• Line no. 2. It occurs only as a weak line in two quasars. Lines at about this wavelength have been known in classical W-R spectra for many years. We divide these lines in three groups according to their wavelengths. Some of the reported wavelengths and assignments of lines in these three groups are: (i) 4577, 4581 Å, Si III (Bears, 1930), 4572.8 Å, Si III (Swings, 1942), 4575.2 Å, O III, O IV (Underhill, 1959), (ii) 4585.2 Å, C III (Underhill, 1959), (iii) 4592 Å (Campbell, 1894), 4593 Å (Wright, 1918), 4593 Å (Bears, 1930), 4593.6 Å, C III (Underhill, 1962).
• Line no. 3. He II 4686 and another line, which very likely is C IV λ4658. We have noted earlier the large variations in the intensity of the latter line. It seems that in most of the quasars considered here the contribution due to C IV 4658 is not very significant.
• Line no. 4. A weak line at about 5090 Å occurs in Sand 4, which Freeman et al. (1968) have attributed to C IV 5093. In classical W-R spectra, Swings (1942) has reported a line at 5092.9 Å, and Underhill (1959) at 5090.8 Å. Underhill (1959) also attributes this line to C IV. However, Moore (1970) does not list any line at about this wavelength for C IV.
• Line no. 5. There is a strong line in NGC 6905 and a medium strength line in NGC 7026 at about 5292 Å, which Aller (1976) identifies with O VI 5292. There is a strong, wide line at 5293 Å in Sand 4. Freeman et al. (1968) consider the contributors to be C III 5273, O IV 5305, and O VI 5291. A line at about this wavelength has been known in W-R spectra since the work of Wright (1918), some of the reported wavelengths and assignments are: 5305 Å (Wright, 1918), 5304.67 Å, O IV (Swings, 1942), 5303.7 Å, C III, O IV (Underhill, 1959).
• Line no. 6. Swings (1942) has reported a line at 5650.8 Å, and Underhill (1959), at 5651.3 Å in W-R spectra. The emitter is believed to be C II.
• Line no. 7. A tracing for 3C 48 given by Greenstein and Schmidt (1964) shows this line as a bump in the shoulder of the strong night sky line 5893. It seems that the wavelength and other characteristics of this line are not well determined, being dependent on the accurate subtraction of the night sky line contribution. We have noted earlier that while Burbidge (1968) has reported this line in the spectrum of 4C 37.43, Schmidt (1974) has not.
• Line no. 8. It has been reported in only one quasar, 3C 48. Underhill (1959) has reported a weak line at 6350.7 Å in the W-R star HD 192103, and she attributes it to C III.
• Line no. 9. A weak line in W-R spectra has been reported at 6632 Å and attributed to O IV (Meinel et al., 1975).
• Line no. 10. In W-R spectra a blend of lines due to C II, multiplet 14, having a strong component at 6784 Å has been reported. There is also a line at 6804 Å, which has been attributed to C III (Meinel et al., 1975).
• Line no. 11. Lines at 6857, 6863 and 6872 Å due to C III, multiplet 19, have been observed in the W-R spectra (Meinel et al., 1975).

Some interesting differences between the spectra of some of the objects considered here may be noted. Hα appears with great strength in NGC 7026 but is completely absent in A 78 (Greenstein and Minkowski, 1964). This is obviously connected with the fact that many of the planetary nuclei are known to be deficient in hydrogen. Sand 4 also does not show Hα (there is only a small spike on the tracing of the spectrum due to the nebula). The blend of C IV 5801, 5812 appears with great strength in NGC 6905, NGC 7026 and Sand 4, but none of the ten quasars appear to have this. It may be noted, however, that some of the quasars have not been investigated in this region. Also, these lines are not far from the night sky line 5893 and could have been confused with it.

From our discussion of the spectra of O VI sequence planetary nuclei, Sanduleak stars, and quasars we are led to infer that these are related objects. However, their family relationship is far from clear. As far as the spectra are concerned, Sand 1, Sand 2, Sand 3, and Sand 5 appear to be indistinguishable from quasars. Sand 4 has associated with it what Johnson (1975) calls a 'counterfeit' supernova remnant. Some questions naturally present themselves. Is there a genetic relationship between quasars and planetary nuclei ? Is a quasar an evolutionary stage of a star preceding the formation of a nebula or does it represent the stage when the nebula has completely dissipated? It is worth noting that 3C 48 has a nebulosity associated with it, which would appear to suggest that perhaps the first possibility is more likely. Could it be that some quasars may have extended nebulae (or perhaps H II regions) of such low surface brightness that they are undetectable? We note that A 30 and A 78 have just such extended and faint nebulae (Greenstein and Minkowski, 1964; Abell, 1966).

It is known that planetary nuclei show emission-line spectra in many cases, and these are of varied types, with certain features having no equivalent among spectra of Population I stars (Vorontsov-Velyaminov, 1962; Smith and Aller, 1969). In view of what we have said earlier, it is not surprising to find that quasars show such richness of variety in their spectra.

5. Other Properties

5.1. Magnitudes

When we consider the two stars Sand 1 and Sand 2, we are, however, faced with a puzzle. The apparent magnitudes of these two stars are similar to those in category A, but if we consider absolute magnitudes (which are calculated assuming that these are associated with the Small and Large Magellanic Clouds, respectively) we find that Sand 1 and Sand 2 are brighter than the brightest category A star by 7.7 and 5.3 magnitudes, respectively. These figures are changed to 5.7 and 3.4 magnitudes if space absorption is allowed for. Two possibilities exist; we shall discuss the pros and cons of both.

1. These are indeed bright objects. The example of BD+30°3639 and HD 164270 quoted earlier will support this hypothesis. One may raise the question as to why such objects have not been found in our galaxy. It is possible that Sand 3, Sand 4, and Sand 5 belong to this category, but their distances are not known.
2. The possibility exists that Sand 1 and Sand 2 might be foreground stars and not belong to the Small and Large Magellanic Clouds, respectively, as radial- velocity confirmation is lacking (Sanduleak, 1971). In that case these stars can be expected to have the same sort of distances as category A stars and consequently their absolute magnitudes will also lie in the same range of values as those of category A stars.

5.2. Distribution in Galactic Coordinates

If we exclude stars in categories A and B which lie between b" = +10° and -10°, we find that the angular distribution of the ten quasars in category C is not unlike these stars. This exception of the zone between b" = +10° and -10° has to be considered in the broader context of quasars in general.

In Figure 2 we show the distribution of about 400 quasars in galactic coordinates. It will be noticed that only two quasars have been reported between b" = +10° and -10°, and in general there is a scarcity of quasars at low galactic latitudes. We believe the reasons for this sparseness are the following. A strong belief amongst quasar investigators that these are extra-galactic objects and so they should be searched for outside the 'zone of avoidance' for galaxies, which is roughly between b" = +10° and -10°. Some of the existing catalogs which are used as a starting point for quasar searches are confined to high galactic latitude objects. For example, LB and PHL catalogs are especially devoted to faint blue stars at high galactic latitudes. Similarly, the Parkes catalog excludes the regions near the galactic plane, except those between l" ~ 165° and 210°.

It is known that there is a moderate crowding of planetaries towards the plane of the Galaxy (Perek and Kohoutek, 1967). From the similarities that we have drawn between planetary nuclei and quasars, we are led to expect that quasars should also occur near the galactic plane.

The average apparent magnitude of the twelve planetary nuclei in Table II is 14.5, and that of the nine quasars is 17.4. Thus the number of quasars that we can expect to observe in the zone b" = +10° to -10° will not be very large.

An underexposed spectrogram (which will suppress the nebular lines) of an O VI sequence planetary nucleus whose spectrum exhibits broad emission lines will be difficult to distinguish from a quasar spectrum. Indeed, as far as the spectra of stars like Sand 1, Sand 2, Sand 3 and Sand 5 are concerned, these stars can be considered as quasars, and can be assigned 'red shifts' of ~ 1.46 or ~ 1.015 following the usual procedures of red shift determination (Schmidt, 1965). Sand 4 exhibits very faint nebular lines and if we ignore them, its spectrum becomes just like a quasar spectrum. Sand 5 has b" = 0.3° and it can be considered to belong to the population of quasars in the zone b" = +10° to -10° that we have predicted earlier.

5.3. Temperatures, Masses

The masses of most of the planetary nuclei are believed to lie in the range of about 1 to 1.5 M_o, though some might have their mass as low as 0.85 M_o and some as high as 4 M_o (Salpeter, 1971). In general, similar values can be anticipated for the masses of quasars.

6. Predictions

1. Most of the quasars in our list should show the resonance line He II λ303.8.
2. The line He II λ3203 should be observable (though weak as compared to He II 4686) in some of the quasars considered in this paper. The reasons are as follows. We have carried out calculations (similar to those of Varshni and Lam, 1976) for the occurrence of laser action in He II λ3203. The results suggest that over certain regions in the ne, Te diagram, laser action in He II λ4686 should be accompanied with laser action in He II λ3203. Also, the central stars of NGC 6905 and NGC 7026 both show He II λ3203 (Aller, 1968).
3. A good number of quasars should be observable in the zone b" = +10° to -10°.

Acknowledgements

Note added in proof. It is of interest to note that recently A. Stockton (1976, Astrophys. J. Letters 205, L113) has reported the existence of a nebulosity surrounding the quasar 4C 37.43. Also, the distribution of the nebulosity around 4C 37.43 is similar to that of the nebulosity around 3C 48, both in extent and in general appearance.

References

• Abell, G. O.: 1966, Astrophys. J. 144, 259.
• Allen, C. W.: 1973, Astrophysical Quantities, The Athlone Press, London.
• Aller, L. H.: 1943, Astrophys. J. 97, 135.
• Aller, L. H.: 1948, Astrophys. J. 108, 462.
• Aller, L. H.: 1968, in D. E. Osterbrock and C. R. O'Dell (eds.), Planetary Nebulae, IAU Symp. 34, 339.
• Aller, L. H.: 1976, Mem. Soc. Roy. Sci. Liege Ser. 6, 9, 271.
• Aller, L. H. and Faulkner, D. J.: 1964, Astrophys. J. 140, 167.
• Beals, C. S.: 1930, Publ. Dominion Astrophys. Obs. Victoria 4, 271.
• Blanco, V., Kunkel, W., Hiltner, W. A., Chodil, G., Mark, H., Rodrigues, R., Seward, F., and Swift, C. D.: 1968a, Astrophys. J. Letters 152, L135.
• Blanco, V., Kunkel, W., Hiltner, W. A., Lynga, G., Bradt, H., Clark, G., Naranan, S., Rappaport, S., and Spada, G.: 1968b, Astrophys. J. 152, 1015.
• Blanco, V., Kunkel, W., and Hiltner, W. A.: 1968c, Astrophys. J. Letters 152, L137.
• Browne, I. W. A., Savage, A., and Bolton, J. G.: 1975, Monthly Notices Roy. Astron. Soc. 173, 87P.
• Burbidge, E. M.: 1968, Astrophys. J. Letters 154, L109.
• Burbidge, E. M.: 1970, Astrophys. J. Letters 160, L33.
• Burbidge, E. M. and Strittmatter, P. A.: 1972, Astrophys. J. Letters 174, L57.
• Burbidge, G. and Burbidge, M.: 1967, Quasi-Stellar Objects, W. H. Freeman and Co., San Francisco.
• Cahn, J. H. and Kaler, J. B.: 1971, Astrophys. J. Suppl. 22, 319.
• Campbell, W. W.: 1894, Astronomy and Astrophysics 13, 448.
• Cannon, A. J.: 1933, Harvard Obs. Bull. No. 891.
• Capriotti, E. G. and Kovach, W. S.: 1967, Astrophys. J. 151, 991.
• Cohen, M. and Barlow, M. 1.: 1974, Astrophys. J. 193, 401.
• De Veny, J. B., Osborn, W. H., and Janes, K.: 1971, Publ. Astron. Soc. Pacific 83, 611 and a supplement to it.
• Edlen, B.: 1932, Observatory 55, 115.
• Edlen, B.: 1933, Z. Astrophys. 7, 378.
• Edlen, B.: 1956, Vistas in Astronomy 2, 1456.
• Ford, Jr., W. K. and Rubin, V. C.: 1965, Astrophys. J. 142, 1303.
• Freeman, K. C., Rodgers, A. W., and Lynga, G.: 1968, Nature 219, 251.
• Greenstein, J. L. and Matthews, T. A.: 1963, Nature 197, 1041.
• Greenstein, J. L. and Minkowski, R.: 1964, Astrophys. J. 140, 1601.
• Greenstein, J. L. and Munch, G.: 1961, Ann. Rept. Dir. Mt. Wilson and Palomar Obs., Carnegie Inst. of Washington Year Book, p. 80.
• Greenstein, J. L. and Schmidt, M.: 1964, Astrophys. J. 140, 1.
• Heap, S. R.: 1975, Astrophys. J. 196, 195.
• Johnson, H. M.: 1973, Mem. Soc. Roy. Sci. Liege Ser. 6, 5, 121.
• Johnson, H. M.: 1975, Astrophys. J. 198, 111.
• Johnson, H. M. and Golson, J. C.: 1968, Astrophys. J. Letters 154, L77.
• Kinman, T. D. and Burbidge, E. M.: 1967, Astrophys. J. Letters 148, L59.
• Kuhi, L. V.: 1968, in K. B. Gebbie and R. N. Thomas (eds.), Wolf-Rayet Stars, p. 103, N.B.S. Special
• Publication 307, U.S. Govt. Printing Office, Washington.
• Lang, K. R.: 1974, Astrophysical Formulae, Springer-Verlag, New York.
• Livingston, W. C.: 1967, Advances in Electronics and Electron Physics 23, 347.
• Livingston, W. C., Lynds, C. R., and Doe, L. A.: 1966, Advances in Electronics and Electron Physics 22B, 705
• Lynds, C. R.: 1967, Astrophys. J. 147, 837.
• Lynds, R. and Wills, D.: 1972, Astrophys. J. 172, 531.
• Matthews, T. A. and Sandage, A. R.: 1963, Astrophys. J. 138, 30.
• Meinel, A. B., Aveni, A. F., and Stockton, M. W.: 1975, Catalog of Emission Lines in Astrophysical Objects, The University of Arizona, Tucson.
• Moore, C. E.: 1970, Selected Tables ol Atomic Spectra, NSRDS-NBS3, Section 3, U.S. Govt. Printing Office, Washington.
• Payne, C. H.: 1933, Z. Astrophys. 7, 1.
• Plaskett, J. S.: 1924, Publ. Dominion Astrophys. Obs. Victoria 2, 346.
• Perek, L. and Kohoutek, L.: 1967, Catalog of Galactic Planetary Nebulae, Czechoslovak Academy of Science, Prague.
• Robinson, L. B. and Wampler, E. J.: 1972, Publ. Astron. Soc. Pacific 84, 161.
• Roeder, R. C. and Dyer, C. C.: 1972, Nature Phys. Sci. 235, 3.
• Salpeter, E. E.: 1971, Ann. Rev. Astron. Astrophys. 9, 127.
• Sanduleak, N.: 1968, Publ. Astron. Soc. Pacific 80, 470.
• Sanduleak, N.: 1969, Contrib. Cerro Tololo Inter-American Observatory, No. 89.
• Sanduleak, N.: 1971, Astrophys. J. Letters 164, L71.
• Schmidt, M.: 1965, Astrophys. J. 141, 1295.
• Schmidt, M.: 1966, Astrophys. J. 144, 443.
• Schmidt, M.: 1974, Astrophys. J. 193, 509.
• Seaton, M. J.: 1966, Monthly Notices Roy. Astron. Soc. 132, 113.
• Smith, H. J.: 1955, Ph.D. Thesis, Harvard University, Cambridge.
• Smith, L. F. and Aller, L. H.: 1969, Astrophys. J. 157, 1245.
• Smith, L. F. and Aller, L. H.: 1971, Astrophys. J. 164, 275.
• Strittmatter, P. A., Carswell, R. F., Gilbert, G., and Burbidge, E. M.: 1974, Astrophys. J. 190, 509.
• Swings, P.: 1942, Astrophys. J. 95, 112.
• Underhill, A. B.: 1959, Publ. Dominion Astrophys. Obs. Victoria 11, 209.
• Underhill, A. B.: 1962, Astrophys. J. 136, 14.
• Varshni, Y. P.: 1973, Bull. Am. Phys. Soc. 18, 1384.
• Varshni, Y. P.: 1974a, Bull. Am. Astron. Soc. 6, 213.
• Varshni, Y. P.: 1974b, Bull. Am. Astron. Soc. 6, 308.
• Varshni, Y. P.: 1974c, Astrophys. J. Letters 193, L5.
• Varshni, Y. P.: 1975a, Astrophys. Space Sci. 37, Ll.
• Varshni, Y. P.: 1975b, Astrophys. J. 201, 547.
• Varshni, Y. P.: 1976a, Astrophys. Space Sci. 42, 369.
• Varshni, Y. P.: 1976b, Astrophys. Space Sci. 43, 3.
• Varshni, Y. P. and Lam, C. S.: 1976, Astrophys. Space Sci. 45, 87.
• Vorontsov-Velyaminov, B. A.: 1962, Comm. Sternberg Astron. Inst. 118, 3.
• Wampler, E. J.: 1974, in N. Carleton (ed.), Methods of Experimental Physics, Vol. 12: Astrophysics, p. 237, Academic Press, New York.
• Wampler, E. J., Robinson, L. B., Burbidge, E. M., and Baldwin, J. A.: 1975, Astrophys. J. Letters 198, L49.
• Westerlund, B. E.: 1964, Monthly Notices Roy Astron. Soc. 127, 429.
• Wills, D. and Wills, B. J.: 1974, Astrophys. J. 190, 271.
• Wright, W. H.: 1918, Publ. Lick Obs. 13, Part VI.

• The GSC object/position is for the hub of the bright bar that looks roughly like a late-type barred spiral galaxy, which is slightly north of the geometric centre. -- Brian A. Skiff

Laser Stars Home