The interpretation of early-type spectra

In this paper the present position concerning the meaning of spectral type among the O- and B-stars in terms of physical conditions in the stellar atmosphere and the abundances of the elements is reviewed. In Section 2 the question “what is a spectral type” is examined and it is noted that because absorption-line spectral types and photometrical spectral types are based on different criteria, the one is not necessarily always consistent with the other. The slope of the Paschen continuum appears to be a good index of effective temperature at types B2 and later, while the wings of Hγ give a consistent means of determining log g. In Section 3 the meaning of spectral type is examined from a theoretical viewpoint. First the simple physical concepts underlying present theories for constructing model atmospheres and predicting theoretical spectra are reviewed and then the results obtained from a series of model atmospheres representing, in principle, main-sequence stars of types B6 to O9 are compared with observed details in stellar spectra. It is shown that the adopted simple theory of spectrum formation, which is based on the concept of local thermodynamic equilibrium, is completely inadequate for understanding the formation and strengths of the strong lines which are used to define spectral types B1 to O5, and even among the stars of type B2 and later the agreement between predicted and observed spectra leaves much to be desired. Examples are presented to demonstrate the difficulty of interpreting line shapes in a unique manner. In Section 4 it is shown that the electron density in the parts of the atmosphere where the classification lines are formed is low enough that considerable doubt exists about the validity of the concept local thermodynamic equilibrium for many key lines and a summary is given of the observational evidence indicating that many of the dominant characteristics of O- and B-type spectra are due to particular processes of excitation and ionization