We present preliminary results about the magnetic field of the Ap star HD 125248, from spectra recorded in RCP and LCP light with the Zeeman analyzer of the CASPEC at ESO.

Spectra and magnetic measurements of the strongly magnetic Ap spectrum variable 53 Cam have been modelled using a line synthesis programme that calculates LTE line profiles, including magnetic and polarization effects and non-uniform distribution of elements over the star. The magnetic field is found to be reasonably well modelled with colinear dipole, quadrupole and octupole components of strengths (at the strong negative magnetic pole) of −16.3, −7.3, and +4.9 kG respectively. Chromium and iron are almost uniformly distributed with abundances of about 102 and 10 times solar. Ti and Ca, in contrast, are quite non-uniformly distributed; Ti is more than 10 times overabundant near the negative pole and probably underabundant elsewhere, while Ca is overabundant around the positive pole and slightly underabundant elsewhere.

Magnetic field diagnosis in three late-type stars is performed using the multiline technique originally developed by Stenflo and Lindegren (1977) to study spatially unresolved magnetic features on the sun.

SiII lines of magnetic Bp-Si stars in open clusters have been observed with the CAT (ESO) in order to get a mapping of the Silicon abundance distribution over the stellar surface, in the frame of the oblique rotator model. We point out the influence of the Zeeman splitting and of the abundance inhomogeneities on the line profiles.

The fundamental parameters and the physical modelling and data used in the analysis of high-resolution high S/N spectra of stars are discussed. Particular emphasis is led on recent developments in these respects and the importance of further improvements is stressed.

The relationship between the r.m.s. photometric accuracy on the data points of a stellar spectrum and the final accuracy on abundances and physical parameters derived from these data is discussed.

We developed routines for intensity, equivalent width, and radial velocity measurements on CES plus reticon spectra obtained with the ESO 1.4m telescope. In order to achieve the optimal recovery of the signal, the noise has to be minimized by removing any parasitical effect. Special care has been devoted to the correction for remanence effects in the reticon dark counts. Typical results are presented and discussed.

We discuss the three questions: Why do we need accurate atomic data for stellar spectroscopy? What accuracy is needed? And: How can the accuracy of oscillator strengths be assessed? In conclusion we comment on the state of the art, stress the importance of uncertainty estimates and also discern between precision and accuracy.

Most of modern solar physics is not concerned with high S/N spectrometry, but rather with high spatial resolution in narrow-band imaging as well as spectrometry, and with high Fourier resolution in helioseismology.

A careful intercomparison is made of several sets of high S/N spectrsocopic observations made by different astronomers, using different telescopes and detectors, observing the same objects. Spectra of moderately weak (10-50 mÅ) Li lines, observed at S/N of one to several hundred using the CFHT Reticon, Lick Observatory Reticon, McDonald Observatory Digicon, and the Intermediate Dispersion Spectrograph plus CCD of the Isaac Newton Telescope agree within the claimed measurement accuracies of about 3-6 mÅ almost without exception. CFHT Reticon observations of Hyades dwarfs are compared with unpublished observations taken with a CCD at Lick Observatory; again, the agreement is very good. However, abundance determinations based on these measurements agree less well. The discrepancy arises from different temperatures used by different investigators for the same stars. The accuracy of various methods of temperature determination is examined, including B-V, R-I, and V-K colors, continuum scans, Hα line profile fitting, and temperatures derived from fine analyses. With careful work and good data, temperatures still limit the accuracy of some of the programs discussed.

The study of lithium depletion in solar-type stars should help elucidate essential properties of convection in those objects. Recent high-quality observations have revealed extensive flaws in our understanding of this phenomenon. The state of our knowledge of Li depletion is reviewed, with a brief possible explanation of it all.

Logarithmic surface gravities between 2.9 and 3.8 (cgs) have been determined for eight G8IV-K2III stars from the analysis of pressure broadened wings of strong metal lines. Comparisons with trigonometrically determined surface gravities give support to the spectroscopic results. A thorough and rather conservative error analysis suggests that the errors are smaller than about 30% and confirms that the method is very insensitive to uncertainties in the fundamental stellar atmospheric parameters. Surface gravities of α Cen A and B and of Arcturus have also been obtained. Chemical abundances have been derived for 12 elements from Na to Ni, and for five heavier elements from Y to Nd have tentative abundances been derived from a single line of each element. Effective temperatures in the photometric system of Frisk (1983) are found to be consistent with spectroscopic temperatures from the excitation equilibrium of Fe I. Surface gravities determined from the ionization equilibria of Fe and Si are found to be systematically lower than the strong line gravities, which may be an effect of errors in the model atmospheres, or departures from LTE in the ionization equilibria.

Spectroscopic analyses of hot pre-white dwarfs, i.e. subluminous O and B stars, are presented. In the B-type stars, the resulting abundance patterns are indicative of atmospheric diffusion (gravitational settling). Amongst the O-type subdwarfs, a new group of comparatively luminous stars is identified. Their position in the HR-diagram suggests that, unlike the “classical” sdOs, they are in a post-AGB stage of evolution. Spectroscopic evidence is presented showing that the born-again post-AGB star scenario of Iben et al. (1983) can explain their origin.

A non-LTE abundance determination of magnesium, iron, and barium was carried out for the atmosphere of Vega. The results indicate marked underabundances of these elements relative to the solar values. First preliminary results of numerical simulations of convective and oscillatory phenomena in the atmosphere of Vega are briefly described.

High resolution, high signal-to-noise observations are required for a better understanding of the Am stars and related topics. Two examples are shown : the relationship between lithium content and temperature in the Am stars of the Hyades cluster and the observational problem induced by multiple spectroscopic binarity frequent among the Am stars.

We have studied the spectra of UU Her stars 89 Her and HD 161796 in the red spectral region. The atmospheric parameters are derived using model atmospheres and spectrum synthesis. LiI line at 6707.8 A° is identified and measured in 89 Her leading to log (Li/H) = 2.3. The variation of the Hα line profile for these stars is also discussed.

Elemental abundances in the Sun and in meteorites are examined with respect to their qualification as a standard of population I composition. Some problems that arise in connection with spectroscopic determinations of abundances in the Sun, and in other late-type stars, are discussed with emphasis on photospheric models and departures from LTE.

High resolution, high S:N spectra are used to determine the abundances of Fe, Ni, Ca, Al and Si in 25 field dwarfs with −1.2<[Fe/H]< +0.3. We find overabundances for Al, Ca and Si in stars with −1.2<[Fe/H]< −0.5 and solar [Ni/Fe] over the whole studied range.

The high resolution and the high light-gathering power of a Fourier–transform spectrometer (FTS) afford the observation of very weak lines in laboratory spectra. Thus it became possible to determine an accurate solar iron abundance from Fe II lines that are weak in the solar spectrum: we measured the branching fractions of a few such lines whose upper levels lifetimes are known.

Four solar type G stars claimed to be photometrically very similar to the Sun have been analyzed in detail on high resolution, high S/N spectra. Their atmospheric parameters : effective temperature, spectroscopic gravity, microturbulence and iron abundance, [Fe/H], have been determined.