A continuous star formation rate in dwarf galaxies?

Various observations of dwarf starburst galaxies reveal homogeneous abundance within the ionized region and, in the case of IZw 18, also within the HI halo. These results seem in contradiction with the models assuming that the measured metallicities result from the metals ejected during the current burst because they should not have time to mix homogeneously at such large scale. It then appears that star formation has occured previously and that these dwarf starburst galaxies are not truly young. On the basis of observational evidence, it is proposed that the metallicity observed in these objects can result only from a low but continuous star formation rate during the galaxy lifetime.