We present chemical abundance determinations of two H II regions in the dIrr galaxy Leo A, from GTC OSIRIS long-slit spectra. Both H II regions are of low excitation and seem to be ionised by stars later than O8V spectral type. In one of the H II regions we used the direct method: O$^{+2}$ ionic abundance was calculated using an electronic temperature determined from the [O III] $\lambda\lambda$4363/5007 line ratio; ionic abundances of O$^+$, N$^+$, and S$^+$ were calculated using a temperature derived from a parameterised formula. O, N and S total abundances were calculated using Ionisation Correction Factors from the literature for each element. Chemical abundances using strong-line methods were also determined, with similar results. For the second H II region, no electron temperature was determined thus the direct method cannot be used. We computed photoionisation structure models for both H II regions in order to determine their chemical composition from the best-fitted models. It is confirmed that Leo A in a very low metallicity galaxy, with 12+log(O/H)=7.4$\pm$0.2, log(N/O)=$-$1.6, and log(S/O)=$-$1.1. Emission lines of the only PN detected in Leo A were reanalysed and a photoionisation model was computed. This PN shows 12+log(O/H) very similar to the ones of the H II regions and a low N abundance, although its log(N/O) ratio is much larger than the values of the H II regions. Its central star seems to have had an initial mass lower than 2 M$_\odot$.

The kinematics of the plasma in 14 PNe is analysed by measuring the expansion velocities (V$_{exp}$) of different ions as derived from their collisionally excited lines (CELs) and optical recombination lines (ORLs). V$_{exp}$ are analysed as a function of the ionisation potential of ions that at first approximation represents the distance of the ion to the central star. In most cases the kinematics of ORLs is incompatible with the kinematics of CELs at the same ionisation potential, specially if CELs and ORLs of the same ion are considered. In general V$_{exp}$ from ORLs is lower than V$_{exp}$ from CELs indicating that, if the gas is in ionisation equilibrium, ORLs are emitted by a gas located closer to the central star. The velocity field derived from CELs shows a gradient accelerating outwards as predicted from hydrodynamic modelling of PNe ionisation structures. The velocity field derived from ORLs is different, in many cases the velocity gradient is flatter or nonexistent and high and low ionised species present nearly the same V$_{exp}$. In addition, the FWHM(ORLs) is usually smaller that the FWHM(CELs). Our interpretation is that ORLs are mainly emitted by a plasma that coexists with the plasma emitting CELs, but does not fit the ionisation structures predicted by models. Such a plasma should have been ejected in a different event that the plasma emitting CELs.

In Planetary Nebulae (PNe) and HII regions ionic abundances can be derived by using collisionally excited lines (CELs) or recombination lines (ORLs). Such abundances do not coincide for the same ion and usually abundances from ORLs are larger than those from CELs by factors of 2 or larger. The origin of the discrepancy, known as the Abundance Discrepancy Factor is an open problem in astrophysics of gaseous nebulae. It has been attributed to temperature fluctuations in the plasma, tiny metal-rich inclusions embedded in the H-rich plasma, gas inhomogeneities or other processes. In this work we analyze the kinematical behavior of CELs and ORLs in two PNe ionized by [WC] stars, finding that kinematics of ORLS is incompatible with the kinematics of CELs. In particular the expansion velocities from CELs and ORLs for the same ion are different, indicating that ORLs seem to be produced in zones nearer the central star than CELs. This is in agreement with results found by other authors for individual PNe.