The Edinburgh spectrophotometric programme aims at the measurement of equivalent widths of stellar spectral lines with the highest accuracy reasonably attainable using the available equipment. In this paper a general description is given of the technique which has been developed; it supplements the more detailed description previously published. All the lines measured lie in the spectral range 4100 A to 6800 A over which tolerable focus can be secured with the spectrograph used. For each star the equivalent widths given depend on the mean of five or six negatives and each negative has been measured at about 1500 fixed wave-lengths. Experience has confirmed that the methods used achieve a substantial gain in accuracy. Systematic errors in calibration have been reduced to the order of one per cent. Plate grain fluctuations have been reduced by combining the measures for the several negatives of each star. A further reduction in accidental fluctuations can be secured by averaging for several stars of the same type, and the formation of a mean plot for five Br stars (27 negatives), when combined with the information given in Moore's Tables, has made it easier to choose continuous spectrum wave-lengths clear of weak lines with central depressions of the order of one per cent. In this way the falsification of the interpolated continuum which can result from the presence of weak lines is reduced. On the mean B r plot the standard errors of measurement at one wave-length are ± 0.008 and ± 0.006 magnitudes in the blue and red respectively. This mean B1 plot also reveals the presence of faint blending lines and a reasonably satisfactory treatment of blends becomes possible. Equivalent widths of 28 spectral lines are given for each of seven stars. For each line the limits of integration used and the standard error of the equivalent width are also given. The standard error depends, inter alia, on the range of integration; for the weaker lines the s.e. is about ± 0.07 equivalent angstroms. For lines other than hydrogen or helium there is a systematic dependence of equivalent width on stellar luminosity, the equivalent widths being larger for the higher luminosities. This progression covers a range of more than two to one in equivalent width. For the Balmer lines H?, H? and H? the progression is reversed, the equivalent widths being smaller for the higher luminosities. The results for the He 1 lines are of special interest. When the singlet and triplet lines 6678 and 5876 are included in a discussion of the material, it becomes impossible to reconcile the equivalent widths with the theoretical curve of growth. There is a deficiency in the measured equivalent widths of 6678 and 5876 which seems to increase with decreasing luminosity. It is suggested that cyclical transition processes are operative, and are more pronounced for the lower luminosities. The Balmer lines of ? Cephei have been found to be variable, the variation being especially pronounced for H?. From the present material no statement can be made as to whether the variation is periodic or irregular.
- 1 - natural sciences ; 2 - physics & astronomy ; 3 - astronomy & astrophysics
- 1 - sciences appliquees, technologies et medecines ; 2 - sciences biologiques et medicales ; 3 - sciences biologiques fondamentales et appliquees. psychologie
- 1 - Physical Sciences ; 2 - Earth and Planetary Sciences ; 3 - Space and Planetary Science
- 1 - Physical Sciences ; 2 - Physics and Astronomy ; 3 - Astronomy and Astrophysics