...108?

Based on observations collected at the Observatoire de Haute-Provence (France).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...é

Research Fellow FNRS (Belgium).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... Rauw

Postdoctoral Researcher FNRS (Belgium).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...fqk

The "f'' tag stands for N III $\lambda \lambda$4634, 4641 and He II $\lambda$4686 present in emission, the "q'' is for the presence of P Cygni profiles and "k'' for the presence of Ca lines which are of interstellar nature.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... Underhill

The errors quoted in Table 3 of Underhill (1994) are not the standard deviations of the data, contrary to what we used in Figs. 4 or 6, but the standard deviations of the means: those two values differ by a factor $\sqrt{N}$. This explains the rather low value of the $\sigma$ in her table.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... low-resolution

The dispersion was 44 Å/mm^-1.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...keV

This flux is in very good agreement with the 0.032ctss^-1 EINSTEIN-IPC count rate quoted by Chlebowski et al. (1989).

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... sequence

Assuming that the typical initial masses of O3, O6 and O7.5 stars are respectively 85, 30 and $20~M_{\odot}$.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.