Pub. Astron. Soc. Pacific, Volume 83, October 1971 ON

Pub. Astron. Soc. Pacific, Volume 83, October 1971

ON THE PRESENCE OF He3 IN THE PHOTOSPHERE OF RHO LEONIS

GEORGE WALLERSTEIN* Berkeley Astronomy Department, University of California Received 8 June 1971

From measurement of 17 spectrograms we find no evidence for the He3 that was pre- viously reported to be present in p Leonis. The profile of Ha was found to show emission on one occasion and to be extremely broad and shallow at another time. Key words: stellar spectra — isotope abundances

I. Introduction second- and third-order iron arc lines. About a In 1967 F. Gutmann reported the ratio of dozen lines of elements other than hydrogen and He3/He4 in the atmosphere of the B1 Ih star helium were used to establish the radial velocity ρ Leonis to be about 1.5. Such a high abundance of the star for each spectrogram. Most of the of He3 in a normal star is surprising. The only lines are due to Ν π, but one or two lines of star with a confirmed detection of He3 is 3 C π, Al m, or Si m were often included. All Centaurus (Sargent and Jugaku 1961; Rodgers lines in ρ Leo are broadened by a combination and Bell 1964). The discovery of He3 in a normal of rotation and turbulence (Huang and Struve supergiant suggests a number of very unexpected 1953). The helium lines do not show the wings possibilities as to its origin. Prior to expending a that might be expected from Stark effect and great deal of effort to explain the phenomenon it appear to be about as broad, though deeper, seemed best to confirm its presence. than the lines of Ν π. The effective wavelength of each helium line was determined for each 3 4 II. The Ratio of He to He plate using the stellar velocity. Because our re- We have obtained 17 spectrograms of ρ Leo sults differ from those of Gutmann (1967) we with the coudé spectrographs of die Hale Obser- present our data in full in Table I. On some vatories. Three different setups yielding disper- spectrograms either the comparison arc or the sions of 15.2, 13.5, and 6.7 Â/mm were used on stellar spectrum was too weak to permit meas- various occasions over an interval of 16 months. urement of lines near 5000 Â; hence the blank All spectrograms were exposed in the visual-red spaces in Table I. region on 103a-F emulsions extending from In Table II we list the measured mean wave- 4800 Â to 6800 Â, since the lines of helium in this lengths of the five helium lines and the laboratory region provide the largest isotope shifts. Several wavelengths for He4 and for He3 (Fred et al. spectra were purposely exposed rather heavily 1951). Clearly the measured wavelengths agree and lightly (within limits that they be measur- with the He4 wavelengths with a systematic able) to test if any wavelength shifts could be difference with respect to the other lines, largely correlated with exposure density. Furthermore, Ν π, indicating a mean outward motion of 2.8 our spectra show some variety of Ha profile from km/sec of He i. The only indication of He3 is a symmetrical absorption to Ρ Cygni-type emis- mean shift of 0.035 Â of the four singlet lines sion. Thus we can also test whether there is any with respect to λ5875; there is no correlation of correlation of the wavelengths of the helium lines shift with the amount predicted for each line. with the Ha profile. The mean predicted shift is 0.36 Â so the 3 Each spectrogram was measured on an oscillo- presence of 10 percent He is possible. The scope measuring engine with respect to the significance of the mean shift of 0.035 Â of the singlets with respect to λ5875 is doubtful be- ^Regularly at the Astronomy Department, University cause the three spectrograms showing Ρ Cygni of Washington, Seattle. emission at Ha all show a discrepancy of 0.15 Â 664

TABLE I Wavelengths of Helium Lines on Each Spectrogram

PLATE DATE DISPERSION EXPOSURE OBSERVED WAVELENGTHS (ur) (A/mm) λ^921 X5015 \50h7 λ5875 λ6678 %

Pb IOU36 19.3 Feb 1968 6.7 Medium 1.83 5.58 5Λ3 8.10 Ρ Cyg On. 10^37 a 19.3 Feb 1968 6.7 Medium 1.71 5.76 5.kh 8.18 Ρ Cyg Sn. b I9.3 Feb 1968 6.7 Medium 1.90 5.76 7.67 5Λ5 8.06 Ρ Cyg Em. Pc IO536 a 5.2 June 1968 13.5 Medium 5.79 8.29 sym. abs. b 5.2 June 1968 13.5 Light 2.02 5.67 8.23 sym. abs» c 5.2 June 1968 13.5 Light 1.97 5.75 8.U1 sym. abs. Ce 19397 a 29.5 Nov 1968 15.2 Medium 1.83 5.67 7.83 5.67 8.05 broad abs. b 29.5 Nov 1968 15.2 Medium 1.85 5.7^ 7.70 5.63 8.16 broad abs. c 29.5 Nov I968 15.2 Medium 1.87 5.78 7.69 5.63 8.10 broad abs, Ce 19399 a 30.5 Nov 1968 15.2 Heavy 1.78 5.6U 7.68 5.56 8.1U broad abs. b 3O.5 Nov 1968 15.2 Heavy- 1.81 5.63 7.64 5.57 8.03 broad abs. c 3O.5 Nov 1968 15.2 Heavy 1.82 5.51 7.56 5.62 7.99 broad abs. Ce 19^3 a 1.5 Dec 1968 15.2 Light 1.69 5.68 7.78 5.58 S.Ik broad abs, b 1.5 Dec 1968 15.2 Light 1.86 5.63 5.65 8.18 broad abs. c 1.5 Dec 1968 15.2 Light 1.93 5.75 7.77 5.^9 8.05 broad abs, Pb III37 a 3.2 June 1969 6.7 Medium 1.97 5.38 7.7^ 5.52 7.96 sym. abs. b 3.2 June 1969 6.7 Medium 1.97 5.63 l.lh 5.68 8.15 sym, abs.

TABLE II III. The Ha Line in Rho Leonis Mean Wavelengths of Helium Lines Emission at Ha is very rarely seen in Β stars of Measured luminosity class lb; in fact Underbill (1966) has λ 4921.86 5015.65 5047.71 5875.60 6678.13 stated that such stars never show Ha emission. Lab He4 4921.93 5015.68 5047.74 5875.68 6678.18 In Plate I, we present a reproduction of Lab He3 4922.26 5015.89 5048.07 5875.72 6678.68 Pc 10437a of ρ Leo showing Ha emission accom- panied by Ρ Cyg-type absorption. Spectra taken at other phases are also shown. in the wavelength of λ5875, thereby pushing Our spectra cover only four distinct epochs on down the mean wavelength of the line by 0.03 A two of which Ha was a symmetrical absorption and accounting for the entire discrepancy. line. Only in February 1968 did it appear to Probably there is feeble emission on the red have a distinct emission component. On the side of λ5876. We conclude that the ratio of 3 4 nine spectra taken 29 November-1 December He /He in ρ Leo is less than 1:10 and is prob- 1968, the absorption is very broad, and almost ably negligible. Thus once again we find that double. The velocity structure at Ha may be 3 Cen is the only star with a confirmed detection described as follows. of He3. 1. When Ha appears as a symmetrical absorp- We have employed the data in Table I to seek tion line the radial velocity is displaced by —9 correlations between dispersion, exposure km/sec. density, and Ha profile with the wavelengths of the helium lines. Except for the displacement of 2. When Ρ Cyg emission structure is present λ5875 on plates showing Ha emission no correla- the emission line is centered at +33 km/sec and tions could be found and we are at a loss to the change from emission to absorption occurs explain the difference between our measures and at about —26 km/sec. The velocity of the those of Gutmann (1967). absorption line appears to be more negative

4375.93 4388.55 The total width of the velocity structure is not I I much larger than the rotational velocity, 62 km/sec, determined from profiles of weaker lines (Huang and Struve 1963). Further obser- vations are needed to clarify the Ha profiles in „yi , ' *""'T /V'l,"ΐίί'*"γΐ terms of rotation and expansion of the outer 'í£X*<¿A '■ ' " layers of ρ Leo. 29 NOV 1968 IV. Conclusions The primary purpose of this research was to seek confirmation of the presence of He3 in the 19 FEB 1968 photosphere of ρ Leo. We have found no evidence for any He3 at all. While we have pre- sented the strongest case we can that the helium ψ 4 Η α en in ρ Leo is He , we cannot find fault with Gut- mann's (1967) measures. His dispersion was PLATE I perfectly adequate, and he measured as many Spectrograms of ρ Leonis near Η α. Weak emission is plates as we. Only an additional set of measures present in February 1968. The center spectrum appears to be extremely grainy because it was enlarged by 2.3 by a third group can finally settle the problem. times as much as the other two, the original dispersion This research was supported by Grant GP- of which was 6.7 Â/mm. The comparison lines are from the third order. 19573 from the National Science Foundation. REFERENCES than this but cannot be reliably measured, as it Fred, M., Tomkins, F. S., Brody, J. K., and Hamermesh, is very shallow. Possibly the emission is super- M. 1951, Phys. Rev. 82, 406. imposed on a symmetrical absorption line dis- Gutmann, F. 1967, Pub. A.S.P. 79, 5. torting its appearance. Huang, S.-S., and Struve, O. 1953, Ap.J. 118, 463. 3. When the absorption appears to be very Rodgers, A. W., and Bell, R. A. 1964, The Observatory 84, 69. broad, the center of the feature is at —7 km/sec. Sargent, W. L. W,, and Jugaku, J. 1961, Ap. J. 134, 777. On one exposure it appeared to be resolved into Underbill, A. B. 1966, The Early Type Stars (Dordrecht, two components at — 49 and + 34 km/sec. Holland: D. Reidel Publishing Co.), p. 216.