We derived elemental abundances in 27 Cepheids, the great majority situated within a zone of Galactocentric distances ranging from 5 to 7 kpc. One star of our sample, SU Sct, has a Galactocentric distance of about 3 kpc, and thus falls in a poorly in
vestigated region of the inner thin disc. Our new results, combined with data on abundances in the very central part of our Galaxy taken from literature, show that iron, magnesium, silicon, sulfur, calcium and titanium LTE abundance radial distributions, as well as NLTE distribution of oxygen reveal a plateau-like structure or even positive abundance gradient in the region extending from the Galactic center to about 5 kpc.
We have performed a NLTE analysis of the infrared oxygen triplet for a large number of cepheid spectra obtained with the Hobby-Eberly telescope. These data were combined with our previous NLTE results for the stars observed with Max Planck Gesellscha
ft telescope with the aim to investigate oxygen abundance distribution in Galactic thin disc. We find the slope of the radial (O/H) gradient value to be equal -0.058 dex/kpc. Nevertheless, we found that there could be a hint that the distribution might become flatter in the outer parts of the disc. This is also supported by other authors who studied open clusters, planetary nebulae and H II regions. Some mechanisms of flattening are discussed.
(Abridged) Extremely metal-poor stars contain the fossil records of the chemical composition of the early Galaxy. The NLTE profiles of the calcium lines were computed in a sample of 53 extremely metal-poor stars with a modified version of the program
MULTI. With our new model atom we are able to reconcile the abundance of Ca deduced from the Ca I and Ca II lines in Procyon. -We find that [Ca/Fe] = 0.50 $pm$ 0.09 in the early Galaxy, a value slightly higher than the previous LTE estimations. -The scatter of the ratios [X/Ca] is generally smaller than the scatter of the ratio [X/Mg] where X is a light metal (O, Na, Mg, Al, S, and K) with the exception of Al. These scatters cannot be explained by error of measurements, except for oxygen. Surprisingly, the scatter of [X/Fe] is always equal to, or even smaller than, the scatter around the mean value of [X/Ca]. -We note that at low metallicity, the wavelength of the Ca I resonance line is shifted relative to the (weaker) subordinate lines, a signature of the effect of convection. -The Ca abundance deduced from the Ca I resonance line (422.7 nm) is found to be systematically smaller at very low metallicity, than the abundance deduced from the subordinate lines.
