A comparison of the classification systems of luminous stars in the northern Milky Way II and IV, for stars assigned two-dimensional spectral types from objective-prism plates

Abstract

For some years a survey for early-type luminous stars in the northern Milky Way has been underway at the Warner and Swasey Observatory and the Hamburg Observatory, as a joint project by the two institutions (1). With the aid of ultraviolet-transmitting objective prisms (Schott UBK7 glass), it has been possible at both institutions to recognize OB stars at a dispersion of about 600Å/mm. Moreover, the appearance of the Balmer discontinuity at low dispersion affords a valuable aid to two-dimensional classification of stars of late B to late F types (2, 3). These two-dimensional classifications, which are based essentially upon the appearance of the continuum (especially the Balmer discontinuity), the hydrogen lines, the K line, and the G band, are assigned in the MK classification nomenclature since they are well correlated with classifications made by means of the real MK classification criteria. The quality of the correlation depends upon location in the HR diagram, but the random probable error appears to be upward of one MK luminosity class and about a quarter of a full letter division in temperature class, except that luminosity class IV is nowhere well distinguished from class V, and class Iab is not really well distinguished from Ia and Ib.The present Cleveland system of assigning two-dimensional spectral types to stars from ultraviolet-included objective-prism plates tends to assign lower luminosities than those of the LS II (–LS I?) system. This fact alone accounts for certainly most of the difference between the numbers of stars assigned two-dimensional spectral types in LS II and in LS IV. The origin of this systematic difference of over a luminosity class is not well explained, but the difference is hardly alarming. The numbers of OB stars in the two catalogues are quite comparable and this is consistent with the fact that no classification system difference for OB stars is known to exist between these two catalogues.Slit spectrograms so far available support the belief that LS IV agrees better in the mean with the MK system than does LS II, though LS IV may also be slightly overluminous in the mean. There is one factor always tending to confuse the transformation from this objective-prism system to MK spectral types: stars having abnormally strong Balmer discontinuities will usually be included in our catalogues with some kind of MK classification, While those with abnormally weak Balmer discontinuities will more often be excluded or assigned to the OB group. This problem is by no means so serious as to detract significantly from the desirability of making objective-prism surveys that pay special heed to the Balmer discontinuity.