SHELL STARS

Pleione a bright shell star of the Pleiades star cluster in Taurus. Shell stars are often surrounded by large amounts of dust seen as the blue reflection nebulosity in this stellar association. (courtesy Anglo-Australian Obs.)

Gamma Cassiopeia, the brightest shell star in the sky was discovered by father Angelo Secchi in 1866, he described a broad emission line spectrum. This was the second discovery of an emission line star (the first was the nova T CrB).

In 1890 J.Scheiner speculated on the origin of the emission lines:

Occasionally an early type emission-line star may exhibit a series of strong and sharp absorption lines, such a spectrum is described as a shell spectrum; a subclass of Be star defined by Paul Merrill in 1949. Shell absorption spectra have been observed for stars from spectral type A5 III (14 Com) to B0e (Gamma Cas). There is some evidence for a shell of small extent in an O7f star (9 Sge). Merrill and Burwell (1949) have published a partial list of shell stars. Most of the work has been limited to a few shell stars, these include gamma Cas, phi Per, Pleione, Zeta Tau, 17 Lep, Beta1 Mon, 14 Com, 48 Lib, and HD 33232.

Most shell stars have variable lines, only a few of them have stable shells. Absorption lines due to metallic ions like Sc II, Ti II, Mn II, Fe II, Cr II, etc, as well as the Balmer Series are known to occur frequently in the spectra of shell stars. But, in general, shell spectra show a gradation in ionization and excitation which are much more complex than in the reversing layer. In the face of such a daunting classification task, Underhill (1966) states :

The color gradient in the picture below graphically demonstrates the effects of gradation of ionization and excitation as a function of distance from the star.

Color image of 'Ring Nebula' (M57, NGC 6720), the red component corresponds to forbidden [ N II ], green to forbidden [ O III ] and blue to He II. (Photo taken at the Kitt Peak National Observatory, data from Balick (1987), courtesy Bruce Balick)

It was noted by Otto Struve and Swings that the shell always shows a lower degree of excitation than the exciting star. Probably the shell of highest excitation ever observed was that of Gamma Cas during it shell episode of June 1939-October 1940. Baldwin was able to measure hundreds of lines but was only able to identify half of them with Fe III, He I and the Balmer series. Ionization stratification has also been observed in Wolf-Rayet stars.

It is well known that in the spectra of shell stars, absorption lines for which the lower level is metastable, are unusually strong. Struve and Wurm (1938) carried out collisional radiative calculations on neutral Helium and were able to furnish an explanation of this phenomena : Dilution of stellar radiation is quite pronounced in the shell and the Rosseland cycle goes into operation leading to large populations in the metastable levels. This conclusion has been fully substantiated by Goldberg (1941), Wellmann (1952) and Nikitin (1952) by taking into account a larger number of levels in their simulations for the helium atom. It would be safe to assume that the overpopulation of metastable states in the presence of a dilute radiation field is a fairly general phenomenon and, unless there is evidence to the contrary, can be expected to occur for any atom or ion. The far infrared laser recently discovered in the extended atmosphere of the hot radio star MWC 349 takes place among the metastable upper quantum levels of the hydrogen atom.

Struve (1942) presented a very penetrating analysis of the problems regarding shell stars, and raised the following points:

  1. Why do some stars possess tenuous outer atmospheres or shells, while other stars, apparently of identical physical characteristics, do not have such shells ?
  2. What is the origin of a shell and how is it supported, in apparent violation of the laws of mechanics ?
  3. How can we account for the remarkable tendency of nearly all shells to vary either periodically or, more often, in an irregular manner ?
  4. Why do some shells expand, while others are stationary ?
Even today our understanding of these problems is very incomplete.

HELIUM RICH SHELL STARS

As early as 1888 Mrs. Williamina Flemming observed the strange emission line spectrum of upsilon Sagittarii, a star in the southern hemisphere. At Yerkes in the early 1940's Jesse Greenstein a post-doctoral fellow under Struve began working on this star and commented...
Greenstein : ... It had almost no hydrogen. It was made largely of helium, and had much too much nitrogen and neon. It is still a mystery in many ways ... But it was the first star ever analyzed that had a different composition, and I started that area of spectroscopy in the late thirties.
- Jesse Greenstein
Greenstein (1940) estimated a hydrogen to helium ratio of about one thousand times lower than for normal stars. It had extremely weak Balmer lines and no Balmer jump and very strong lines of helium superposed on a complex metallic spectrum. It is believed that there is an extended nebulous envelope around the star. The observed spectrum is that of a single star. This star is a member of a binary system (Wilson, 1914). No evidence in the visible has been detected of the other member, but recent ultraviolet evidence for a companion has been obtained (Parthasarathy et al., 1986). A more recent analysis by Schoenberger and Drilling (1983) indicates a much lower hydrogen abundance, n(H)/n(He) less then 0.0005. Stars with low hydrogen abundance have a low opacity and hence an ultraviolet excess similar to supergiants and many quasars.

Many shell stars occur in clusters because they were all born within the same stellar nursery roughly at the same time. Due to this fact some of the members evolved simultaneously at the same rate, hence their spectra are almost identical such as the stars in the trapezium cluster in Orion nebula or the Pleiades cluster (see picture at top of page). Other stars within the association may have slightly different initial masses or environments; which creates differences in their spectra. In either case the main results is that stars of remarkably identical spectral type or different spectral types but similar ages occur in a particular area on the sky called a stellar association.

COMMENTARY

QUASAR CLUSTERS

Recent findings by Halton Arp (1987) that reveals that quasars with discordant redshifts often occur in tight associations or clusters on the sky. If quasar redshift was a distance indicator, this would translate to the odd picture of a universe filled with long elongated 'fingers' pointing towards the earth. This would mean that the earth is a 'favoured' spot in the cosmos in which all these strange geometrical shapes are pointing towards only us. However, Varshni (1979) has shown that the redshift is an empty number without physical significance: These clusters are nothing more than the stellar associations mentioned above.

MULTIPLE REDSHIFTS

Of historical interest is the fact that while Greenstein was a prominent figure in stellar spectroscopy, he unfortunately started the multiple-redshift hypothesis of quasar absorption lines in 1967. Twenty-eight of the forty-nine absorption lines measured by Greenstein and Sargent (1968) in quasar 0237-233 in the range 3346-4493 Å (in collaboration with Bahcall) were identified with five redshifts systems. However, many of the strong lines were unidentified, even with five redshift systems. It is not without significance that after this work on this quasar, Greenstein, a very experienced stellar spectroscopist, quit quasars altogether and years later wrote (Greenstein, 1984) : (Parenthetically we note that it was not for the lack of telescope time, because as he himself notes, he had been assigned 20 to 30 nights a year (1952-1979) at the 200-inch)

Varshni (1988) has identified 195 absorption lines in quasar 0237-233 with shell star absorption lines from Fe I, Fe II, Cr II, Ni II, Si I, Si II, Mg I, Mg II and Ti II. (and similarly with QSO 0805+046 and 0420-388). This is a clear indication that quasars have atmospheres which are related to hydrogen deficient, helium rich shell stars, further proof that they are stars within our galaxy. Varshni (1981 and 1983) demonstrates that the proposed redshift systems in this quasar are numbers without physical significance.

Web site

References

  1. Arp, H.: 1987, Quasars, Redshifts and Controversies, Berkeley: Interstellar Media. (also Arp's Catalog of Peculiar 'Galaxies')
  2. Bahcall,J.N., Greenstein,J.L., Sargent,W.L.W.: 1968, Astrophys.J., 153, 689.
  3. Baldwin,R,B.: 1941, Astrophys.J., 93, 333. The shell of highest excitation ever observed was that of gamma Cas during its shell episode of June 1939 - October 1940. Out of 249 lines observed only 145 could be identified. Most of the identified lines were due to Fe III, He I and the Balmer series.
  4. Baldwin,R.B.: 1941, Astrophys.J., 93, 420.
  5. Baldwin,R,B.: 1943, Astrophys.J., 97, 388. Detailed study of the systematics of shell spectra. gamma Cas, xi Tau, epsilon Cap, beta Mon, Pleione, HR 8731, psi Per can be arranged in order of decreasing excitation conditions in the shell.
  6. Balick,B.: 1987, Astron.J., 94, 671.
  7. Ballereau,D.: 1980, Astron.Astrophys.Suppl., 41, 305. Atlas of the spectra of Pleione between 3167 and 4924 A (p.315-318), better than one angstrom resolution.
  8. Broyles,A.A.: 1943, Astrophys.J., 97, 234. The spectrum of HD 192954
  9. Fleming,W.P.: 1912, reported by A.J.Cannon, Harvard Ann. 56, 65, 108.
  10. Goldberg,L.: 1941, Astrophys.J., 93, 244. Confirmed the overpopulation of He I metastable states using more level in the collisional readiative calculation.
  11. Greenstein,J.L.: 1974, American Institute of Physics, Oral History Interview by P. Wright, p.17.
  12. Greenstein,J.L.: 1984, Ann.Rev.Astron.Astrophys., 22, 1.
  13. Hack, Struve,O. : 1970, Stellar Spectroscopy II. Peculiar stars, Observatorio Astronomico di Trieste. Review of shell stars including : gamma Cas, Pleione, 48 Lib and xi Tau
  14. Hanuschik,R.W.: 1995, AA., 295, 423. Shell lines in disks around Be stars. 1: Simple approximations for Keplerian disks
  15. Hiltner,W.A.: 1944, Astrophys.J., 99, 103. The shell spectrum of 48 Lib (HD 142983).
  16. Hunger,K.: 1975, in B.Baschek, W.H.Kegel, G.Traving (eds.), Problems in Stellar Atmospheres and Envelopes, Springer-Verlag, New York. Helium rich shell stars.
  17. Merrill,P.W.: 1949, Publ.Astron.Soc.Pacific, 61, 38.
  18. Merrill,P.W., Sanford,R.F. : 1944, Astrophys.J., 100, 14. The spectra of 48 Lib.
  19. Merrill,P.W., Burwell,C.G.: 1949, Astrophys.J., 110, 387.
  20. Merrill,P.W.: 1952, Astrophys. J., 115, 42. Most shell stars show a marked changes in intervals not exceeding a few years in the appearance or position of the lines. Only a few of them have stable shell. (HD 193182, HD 195325, HD 54858)
  21. Merrill,P.W.: 1952, Astrophys. J., 115, 47. Three B-type stars HD 172694, HD 184279, HD 195407 which have sometimes shell spectra marked by absorption lines from the metastable levels of the neutral helium atom and the metallic lines are absent or weak presumably because the excitation required for the helium lines is sufficient to remove a second electron from the metallic atoms. These features appear to indicate that the shells are of high excitation.
  22. Merrill,P.W., Lowen,A.L. : 1953, Astrophys.J., 118, 18. Intercomparison of 21 shell stars. Only a few of them have stable shell. (HD 193182, HD 195325, HD 54858)
  23. Nikitin,A.A.: 1952, Dokl.Akad.Nauk SSSR., 85, 285.
  24. Parthasarathy,M., Cornachin,M., Hack, M.: 1986, Astron.Astrophys., 166, 237.
  25. Russel,H.N., Bowen,I.S. : 1929, ApJ, 69, 196. Chance coincidences in spectral lines.
  26. Scheiner,J.S.: 1890, Spectralanalyse der Gestirne, pub. by W.Engelmann, Leipzig, p.276.
  27. Schoenberner,D., Drilling,J.S.: 1983, Astrophys.J., 268, 225.
  28. Secchi,A.: 1866, Comptes Rendus de l'Academie des Sciences, 63, 621.
  29. Slettebak,A. : 1988, Pub.ASP, 100, 770. The Be Stars.
  30. Struve,O., Roach,F.E.: 1939, Astrophys.J., 90, 727.
  31. Struve,O., Swings,P.: 1941, Astrophys.J., 93, 446. "all known absorption shells form a more or less uniform spectral sequence, which runs parallel to the usual sequence, with the shell nearly always showing a lower degree of ionization than the exciting star"
  32. Struve,O., Wurm,K.: 1938, Astrophys.J., 88, 84. In the shell the dilution of the stellar radiation is quite pronounced and the Rosseland cycle goes into operation leading to large populations in metastable levels. They used a six level collisional radiative model for He . Therefore in the spectra of shell stars absorption lines for which the lower level is metastable, are unusually strong.
  33. Struve,O.: 1942, Astrophy.J., 95, 134. (see text)
  34. Struve,O., Swings,P.: 1943, Astrophys.J., 97, 426.
  35. Struve,O.: 1943, Astrophys.J., 98, 98. The spectrum of 48 Lib (HD 142983)
  36. Underhill,A.B.: 1958, Publ. Dominion Astrophys. Obs. Victoria, 9, 143. Spectra of 9 Sge.
  37. Underhill,A.B. : 1966, The Early Type Stars, D.Reidel Publishing Co., Dordrecht. Spectra of xi Tau and 48 Lib.
  38. Varshni,Y.P.: 1978, 'The Ta-You Wu Festschrift: Science of Matter', ed. S. Fujita (Gordon and Breach, 1978) p.285
  39. Weaver,H.F.: 1952, Astrophys.J. 116, 541. The spectrum of gamma UMi
  40. Wellmanm,Z.: 1952, Z.Astrophysik., 30, 71. Confirmed the overpopulation of He I metastable states using more level in the collisional readiative calculation.
  41. Wilson,R.E.: 1914, Lick Obs. Bull., 8, 134.
Next historical event