2000A&A...361...33N

``Cometary'' Blue Compact Dwarf Galaxies (iI,C BCDs) are characterized by an off-center starburst close to the end of their elongated stellar bodies. This rare phenomenon may carry some clues on how collective star formation ignites and propagates in gas-rich low-mass stellar systems. This off-center burst may be a fortuitous enhancement of the otherwise moderate star-forming activity of a dwarf irregular (dI), or may be caused by a set of special properties of such systems or their environment. We attempt here a first investigation of this issue by analysing two prototypical examples of cometary dwarf galaxies, the nearby iI,C BCDs Markarian 59 and Markarian 71, both containing an extraordinarily luminous HII region in the outskirts of a dI-like host. Using deep ground-based spectrophotometric data, and HST images, we study the physical state of the starburst regions and the structural properties of the underlying irregular galaxies. We find that the average metallicities show small scatter in the vicinity of the star-forming regions and along the major axis of Mkn59 which suggests that mixing of heavy elements must have been efficient on scales of several kpc. The azimuthally averaged radial intensity distributions of the underlying host galaxies in either iI,C BCD can be approximated by an exponential law with a central surface brightness and scale length that is intermediate between typical iE/nE BCDs and dwarf irregulars. Spectral population synthesis models in combination with colour magnitude diagrams and colour profiles yield a most probable formation age of ∼2Gyr for the low surface brightness (LSB) host galaxies in both iI,C BCDs, with upper age limits of ∼4Gyr for Mkn59 and ∼3Gyr for Mkn71, i.e. significantly lower than the typical age of several Gyr derived for the LSB component of iE/nE BCDs. These findings raise the question whether iI,C systems form a distinct physical class within BCDs with respect to the age and structural properties of their hosts, or whether they represent an evolutionary stage connecting young i0 BCDs and ``classical'' iE/nE BCDs. In spite of the scarcity of available data, a review of the properties of analogous objects studied in the local universe and at medium redshifts provides some support for this evolutionary hypothesis.