Most of the suspected WR galaxies come from the spectrophotometric catalogue of HII galaxies of Terlevich et al. (1991), following the analysis of Masegosa et al. (1991) or Pindao et al. (1999) which also includes new objects. As stated earlier, from Masegosa et al. (1991) we take the subset of objects included in their Table 1 but not in Table 2 as candidates. Although partly based on the same observations the analysis of Masegosa et al. (1991) and Pindao et al.(1999) do not always yield the same candidates. We have retained all objects classified as ``candidate'' by either one of these studies. New observations will be necessary to establish the definite presence or absence of WR features.
Few objects from Vacca & Conti (1992) have only an upper limit
on broad HeII 4686. For most of them independent observations are now
available. Otherwise these objects are retained as suspected WR galaxies.
In some studies of IUE spectra strong UV P-Cygni lines of N and C
(N V 1240, C IV
1550) have been interpreted as
signatures of WR stars in theses objects (e.g. Durret et al. 1985, Lamb
et al. 1990).
However, these lines are also strong in O stars and hence cannot be used
as a clear diagnostic for WR stars (cf. Leitherer et al. 1995).
Therefore, objects suspected on these grounds have not been included in
Table
.
In rare cases, WR stars have also been suspected on indirect grounds (e.g. M100: Wozniak et al. 1998).
Interestingly, a broad WR bump is suggested to be present in the optical
spectrum of two distant central cluster galaxies with strong cooling flows
(Abell 1068 and 1835, Allen 1995). The recent study of Contini et al. (1998), however, casts serious doubt about the reality of a broad
feature in Abell 1835.
If true, these objects with redshifts 0.14 and 0.25 respectively,
represent the most distant objects known to date where WR stars have been
detected from (rest-frame) optical spectra.
High redshift galaxies (
3) may, however, also show WR signatures
(see Sect. 7).