The so-called ‘weak helium’ stars are discussed from a spectral classification point of view. Several authors have included different kinds of objects in this class. Often, the only similarity is the discrepancy between the spectra and the colours. It is desirable to decouple the discovery of new members of the class from this dependence on colours so that the colour can be used as an independent parameter.

Comparaison entre notre classification, basée sur les raies de l'hydrogène (Rc γ, δ, ε W γ, δ, ε), et les classifications Chalonge, Barbier-Morguleff et J. Rountree-Lesh. Il semble que notre classification soit voisine de ces trois autres mais qu'elle est plus sûre dans le cas des étoiles Be.

Nous indiquons une liste de 29 étoiles membres du Grand Nuage de Magellan que nous proposons comme standards de type spectral des Supergéantes. La situation de ces étoiles dans les diagrammes RH et couleur-couleur est indiquée. Nous proposons des valeurs de (U − B)0 et (B − V)0 pour ces étoiles.

In December 1970 and January 1971, spectra of 13 LMC supergiants have been obtained with the Chalonge spectrograph attached to the Cassegrain focus of the 152 cm ESO reflector at La Silla. The spectrophotometric parameters λ1, D, φb and φuv were measured and the first results deal with (a) spectral classification and (b) the distance modulus of the LMC.

An investigation has been made of the central region of the cluster IC 2944. For about 70 stars UBV photometry has been made. For the brighter of those stars also Hβ has been measured. Slit spectra of intermediate dispersion (73 Å mm-1) have been made for the 40 brightest stars. Coudé spectra (12 Å mm-1) have been obtained for 8 stars. Fabry-Pérot measurements have been made in 5 points within the nebulosity. The number of blue stars is shown to be extremely high. Down to visual magnitude 11.5 the cluster seems to contain only O and B stars, the majority of spectral class earlier than B3. The Fabry-Pérot measurements give evidence of high internal gas motion. Evidently the gas is thin, visible only because of the great number of hot stars.

Details will be published elsewhere.

Most of the results reported thus far of image tube spectroscopy as applied in astronomy have been concerned with emission features and have been the result of a combination of the largest available light gathering power of large telescopes with the improved sensitivity of image intensifiers. We consider here the application of image tube techniques to moderate sized telescopes; concretely, we describe the first program of spectral classification using absorption features as carried out with the image tube spectrograph of the Vatican Observatory attached to the Zeiss 60-cm reflector at Castel Gandolfo. Treanor (1970) has described the optical design, construction and the first tests of this instrument. The receiver is an RCA cascaded image tube loaned to the Vatican Observatory by the Carnegie Image Tube Committee.

Spectra of 60 M-type long-period and semiregular variables, obtained near the time of maximum at Siding Spring Observatory in Australia, from 1965–1967, have been classified on the Keenan system.

Two hundred giant and subgiant stars near K0 were classified twice by somewhat different techniques. The resulting types were in excellent agreement but a comparison of the assigned types with U, B, V colors indicated both an intrinsic scatter in the colors for stars of similar spectroscopic appearance and a problem in either the luminosity classifications or the standard colors or both. Several peculiar stars and a luminous supergiant were also detected.

Sanduleak and independently Seggewiss discovered a very red star near the open cluster NGC 6231. Bessell and collaborators concluded it to be a variable of the R Coronae Borealis type.

A pair of spectrograms of this object were obtained in July 1970. The spectrograph dispersion in the range of 3800–4500 Å is around 80 Å mm−1 and in the range of 4500–7000 Å 155 Å mm−1.

An outstanding feature of the spectrograms is their ‘veiled’ aspect, making it rather difficult to observe the lines. We observe the 4226 Å of Ca I, H and K of Ca II, 4030/4034 Å of Mn I, the G band; and several lines belonging to the Fe I multiplets. The Hα and Hβ lines are clearly seen in emission. Hδ is absent or very faint. A spectrogram of RY Sagitarii taken in July 16, 1971, when the visual magnitude of the star was around 4 mag. below its maximum light, shows some similarities with the spectrograms of the Sanduleak-Seggewiss star, e.g. they do not show the ultraviolet CN absorption bands and in the spectra of both stars a line is seen at approximately 3896 Å, which may be H8.

If we accept that the star belongs to NGC 6231, its distance modulus is 11.5. Its intrinsic color may be assumed to be close to (B–F)0 = 1.0(G or K star). The observed color is (B–K) = 3.5. Adopting R = 3, the visual absorption is 7.5. As the visual magnitude of the star is V= 12.5, its absolute magnitude becomes – 6.5.

Details will be published elsewhere.

It is shown that a large scatter exists in all color-color and color-spectral type diagrams. This scatter can not be neglected and its importance is shown in several examples. A new approach is then developed for the study of the precision attainable in spectral classification. It is found that for A-type stars MK classification is about as accurate as photometry, whereas for B-type dwarfs photometry is at least two times more precise than MK classification. A precision of ±0.6 tenths of a spectral class seems attainable in both groups.

By examining the observed dispersion in (colour, spectral type) relations, classification errors have been derived from the data of nearby stars. The comparisons of the colour deviations observed in spectral regions of large variations of colour with type with the deviations in regions of small variations give the following standard errors in units of a tenth of a spectral class: For K dwarfs ±0.6 (MK), ±1.2 (Mt. Wilson), ±0.7 (Kuiper); for early M dwarfs ±0.9: (MK), ±0.7 (Mt. Wilson), ±0.5: (Kuiper); and for late M dwarfs ±0.7 (Kuiper).

The measurements of spectral lines of moderate dispersion slit spectrograms for F, G and K type stars give a large quantity of information. The measurements are time-consuming in comparison with visual classification, but the use of digitized microphotometers reduces the task, and the method is now promising. The interpretation of the measurements raises some problems, which are briefly discussed.

Traditionally, introductory talks usually have either the function of justifying the meeting at which they are given, or of attempting to present some sort of overview of the whole subject, or sometimes both or neither. A meeting such as this one is generally in order when the field in question has for any reason been very active since the last meeting about it; and by that standard I should say that this symposium is indeed appropriate inasmuch as the field of spectral classification has surely been very active during the seven years since the last symposium on spectral classification and multicolor photometry. In the area of objective prism spectral classification, the last several years have not seen very much really new in the way of observing techniques or methods of analysis that I know of, at least in comparison with some other fields; but they have certainly seen a great deal of astronomy done.

The major project of assigning spectral and luminosity classes to the Henry Draper stars south of +30° is underway. The high-quality objective-prism plates have been taken with the Michigan Curtis Schmidt telescope at Cerro Tololo Inter-American Observatory. The spectra, widened to .8 mm, have a dispersion of 108 Å mm-1 at Hγ. Plates exposed for 20m and for 4m + 1m yield classifiable spectra between 4 and 10 mpg.

Star identification is completely automatic, using a computer-generated plot to the Schmidt plate scale for each plate center. This plot, having the HD stars identified by number, is copied onto transparent material and placed under the Schmidt plate while the stars are visually classified.

Classification is being carried out by one person (N.H.) to maintain a uniform system. MK spectral standards of similar quality and density, also taken with the Schmidt, are continually referred to. Intercomparison of the new types with existing spectral types shows no systematic differences. However, the HD types themselves are systematically earlier in the range A to F.

Stars between the south celestial pole and δ = −55° are being classified first, with the first volume of results to be published late in 1973. It is expected that the catalogue of spectral types and remarks will be published in 6 volumes containing all HD stars south of +30°.

This research is being supported by the National Science Foundation.

This paper discusses the present status of a preliminary inspection of the plates obtained in the Michigan objective-prism survey of the entire southern sky. A very large number of peculiar stars, and of normal supergiants as well, have been noted and lists of these are being prepared for publication.