We welcome today with special joy the presence among us of scholars who have come from all parts of the world to our small State of Vatican City, to study the Spectral Classification of the Stars. We know that, in the spirit of collaboration, you will confer together on the topics proposed by the International Astronomical Union. We thank you for the honour of your presence and we are glad of this opportunity to show our continued interest in Science and particularly in Astronomy; this interest leads us to maintain the Vatican Observatory and to support our esteemed Academy of Sciences.

A review is given of the present state of MK classification, with a view to future developments in techniques and instrumentation. The principle of the complementarity of quantitative and visual inspection techniques is stressed.

Included in the discussion are examples of problems which are currently outstanding. Among these are variable stars, marginal peculiarities, fundamental standards (with specific reference to the Sun), and representation of the third and higher dimensions.

The fundamental characteristic of the MK classification system is the definition of an array of standard stars, arranged in two dimensions. New techniques and an enormous increase in observational and theoretical work since the original catalog in 191+3 have in various ways complicated the problem of spectral classification. Problems discussed are: 1.) The re-examination of standard stars; 2.) The optimum dispersion and the limits of dispersion for MK classification; and 3.) Some groups of peculiar objects.

The sequence of stars starting with type S and extending through the SC stars into the red carbon stars are brought into the following revised system of classification:

In this table X stands for the temperature type and the number to the right of the slant line is an abundance index which, as suggested by Ake (1978), is essentially a measure of C/O. Intensities of ZrO and TiO are given also.

A brief review and summary of results from Volumes 1 (-90° to -53°) and 2 (-53° to -40°) of the University of Michigan Catalogue of Two-Dimensional Spectral Types for the HP Stars is given. Plans for the classification of the remaining southern HD stars are outlined. Future plans for observation and classification of northern stars are discussed, including the problem of whether to limit such classification to the HD stars. More general remarks about spectral classification in the future, especially for fainter stars, conclude the paper.

The principal domain for the determination of the fundamental reference stars is K - Hβ (3930 - 4860 Å/mm). A certain amount of classification has been carried out in the visual and near ultraviolet regions; however, in neither case has a serious attempt been made to go to the practicable limit in accuracy. It can thus be said that the “field” for the determination of the fundamental MK standards has been K - Hβ.

The definitions and general principles of the MK system given in the statement by Dr. Morgan, which Dr. McCarthy has read to you, are an expression of my views also.

It is mainly in specifying the MK domains that modifications are necessary for the cooler stars - and these for rather obvious reasons. For example, the spectral range has been extended to longer wavelengths, particularly for stars showing strong molecular bands. The lines and bands in the visual and far red regions are particularly useful in defining temperature and chemical composition, but the blue region remains important as having the features most sensitive to luminosity.

Les classes de luminosité de Keenan et Morgan sont une subdivision très commode pour caractériser les étoiles d’une même classe spectrale. Les classes V et III paraissent bien définir un type d’étoiles; leurs paramètres principaux (MB et Te ou pe et Te) sont déterminés par le type spectral et la classe de luminosité. Les classes IV et II sont mal définies, leur existence n’est pas assurée. L’ensemble des étoiles supergéantes est encore plus mal défini, elles remplissent complètement la partie supérieure de diagramme de Russell-Hertzspung. Une étude de ce diagramme pour les étoiles supergéantes du Grand Nuage de Magellan montre que les subdividions IaO, la, lab et Ib correspondent à une variation continue de la magnitude absolue qui va par exemple des étoiles AO de -9.6 à -5. La valeur souvent admise -7 pour la magnitude absolue des supergéantes est une simplification abusive. D’autre part, la classification exacte des luminosités est difficile et une mesure d’une quantité physique (rapport d’intensité de raies, ou intensité de raies) devrait permettre une meilleure classification de ces étoiles.

It has been apparent for some time that there is a need for a single luminosity calibration to be used with modern MK types for early-type stars, at least from 0 through middle B. The widely used calibration of Blaauw (1963) has to be replaced because the refinement of the MK system - as reflected in the large collections of spectral types by Lesh (1968), Hiltner, Garrison, and Schild (1969) and Walborn (1971) - has led to a lower mean luminosity for most main sequence subgroups of early-type stars, as the higher luminosity stars tend to move out of class V. Thus the calibrations of Lesh (1968) and Walborn (1972, 1973) are systematically fainter than Blaauw’s

Stellar rotation may affect the strengths of lines used in spectral classification because of effects of plate resolution and also because of physical changes in the rotating stellar atmospheres. In the first case, the importance of using standard stars with appropriate rotational line broadening in classifying spectra of relatively high dispersion (resolution) is emphasized. Rotational effects on weak lines in general and on the Balmer lines may cause systematic errors in the assignment of spectral types and luminosity classes unless the standard stars are chosen with line broadening similar to that of the star to be classified. With regard to physical changes in the rotating atmospheres, we have extended the earlier work of Collins to include the Balmer lines plus additional lines of He I and Si II which are important in spectral classification over a wider range of spectral types.

Morgan, Keenan, and others have found that low dispersions (80-130 Å mm-1) are ideal for obtaining spectral types and luminosity classes, especially for early-type stars. This is because at those dispersions certain blends are unresolved and remain useful, the hydrogen line wings are seen most clearly, and differences in rotational velocities are not so important. Such dispersions are also ideal or adequate for detecting some peculiarities, such as Am and Be spectra and some extreme Ap spectra, but they are inadequate for detecting abnormalities involving weak lines (as in Ap(Hg, Mn) stars) and line profiles (such as in shell spectra).

Our purpose is to describe the program on spectral classification of members of open clusters and associations started in January 1977 at La Plata Observatory. The program is essentially a continuation of the Abt and Levato studies made from Kitt Peak so, many details of it have been already described in Dr. Abt’s paper in this volume. We have a list of 40 clusters and associations included in the program which consists in obtaining two sets of spectra in two dispersions, 39 Å/mm and 125 Vmm, both sets 1.2 wide. These spectra, as has already been mentioned, are very useful for deriving accurate MK types and for discovering peculiar stars.

The area under discussion here lies beyond the domain of the MK system and includes studies of natural groups as detected in low dispersion studies; it extends to spectral regions studied now in the far ultraviolet and in the infrared and radio regions; in this paper primary consideration is given to optical spectra from the near infrared to the near ultraviolet. We give a description of “natural groups” and a brief description of new possibilities opened for research in this area through the introduction of thin prisms and the combination of prism and grating (GRISM). Finally a summary of spectral criteria useable outside the MK domain is presented.

The use of natural spectral-form groups (Morgan, 1951) on low-dispersion slitless survey plates efficiently isolates rather pure samples of intrinsically interesting extragalactic objects (e.g. Smith, 1975, 1976; Smith, Aguirre and Zemelman 1976; MacAlpine, Smith and Lewis 1977a, b; MacAlpine, Lewis and Smith 1977, MacAlpine and Lewis 1978, Cooke et al. 1977, Hoag and Smith 1977, Sramek and Weedman 1978, Bohuski, Fairall and Weedman 1978, Osmer and Smith 1979). One of these groups (illustrated for various telescopes by Smith 1975, Hoag and Smith 1977, Bolton et al. 1977 and Smith 1978) has been found to include a very large set of high-redshift QSOs. Definition of the group (e.g. Smith 1976) rests on the direct detection of the QSO emission lines on the survey plates. Objects already selected from the group range in redshift from 0.32 (PKS 2227-399 - see also Peterson and Bolton 1978, Browne, Savage and Bolton 1976, Bolton et al. 1977) to 3.45 (Q2227-6-3928 - Smith et al. 1977, Osmer 1978a) and in apparent magnitude from 21 (Osmer 1977, Smith 1978) to 16 (including the apparently brightest high-redshift object known, Q1101-264). - Osmer and Smith 1976, Smith 1978, Carswell et al. 1979 and the most luminous object known, Q0420-388 -Osmer and Smith 1977b; Wright and Kleinmann 1978). 15 of the 21 objects known with confirmed redshifts z > 3 have been discovered this way.