Puppis, happens also to be Of, i.e. it has a relatively strong wind. It is an obvious first target for search of blobs in the winds of OB-stars, if they should exist. We observed Pup at the f/8.2 Coudé focus during the two nights of 10/11 and 12/13 December 1995. Using the red Coudé train and image slicer, the 1800 l/mm holographic grating and the Reticon 1872 array as a detector, we obtained a S/N 
1000/0.03 Å pixel in 10 minutes on a total range of 60 Å centered on HeII 4686 Å. The Reticon 1872 detector has been decomissioned after this run. The data reduction was carried out using IRAF with a Reticon reduction package developed by D. Bohlender and G. Hill, which includes the baseline reduction, flatfielding, heliocentric correction and wavelength calibration with a Thorium-Argon comparison spectrum. The FWHM of the Th-Ar lines covers ~2 pixels and the wavelength shift over the night is negligibly small.
Puppis shows spectral substructures (Fig. 1, and Eversberg, Lépine, Moffat 1996, henceforth ELM) similar to those seen in the winds of Wolf-Rayet stars, which are the likely descendents of Of stars. These substructures are likely the consequence of excess emission from clumps in the wind. We note the following:
(
), follows the line profile itself, with some increase in () on the blue side as seen in WR lines (ELM, Fig. 2). This is compatible with the fact that we detect a similar number of blobs on the 4686 line in Pup as on lines of smaller width in WR stars (e.g., CIII 5696 in WR 103: Robert 1992).
Pup in H
(Moffat & Michaud, 1981)(ELM, Fig. 3). Because of the relatively short coverage (~ 5 hours on each of two nights) we are not able to establish any periodicity in the variation of this absorption component, although sinusodial behavior cannot be excluded. In the case of the near-central absorption in H Moffat & Michaud (1981) found a 5.2-day periodicity, which led them to the model of an oblique rotator for Pup.
` velocity law for Pup with an adopted stellar radius of 18 R* (Kudritzki, Simon & Hamann, 1983) and a terminal velocity of 2250 km/s (Puls et al., 1995), all blobs in HeII 4686 appear near the star's surface and disappear below 2R*. The best `fit' gives = 1.3 (ELM, Fig. 4), although a `' law may not necessarily be strictly valid.The commonly adopted value = 0.8 for OB winds does not appear compatible with our data. As a first approximation we have traced the blobs by straight lines; however, slight non-linear acceleration may also be possible.
Pup is a peculiar O-star, other than its being the apparently brightest O-star in the sky. Thus, it may well be that all OB-star winds show micro-structures as seen in Pup. Of course this must be tested by looking at other bright O stars.
Eversberg, T., Lépine, S., Moffat, A.F.J., 1996 (ELM), in: ``WR Stars in the Framework of Stellar Evolution'', 33rd Liège Int. Astroph. Coll., 1996
Henriksen, R.N., 1994, ApSpSc, 221, 25
Kudritzki, R.-P., Simon, K.P., Hamann, W.-R., 1983, A&A, 118, 245
Moffat, A.F.J., Michaud, G., 1981, ApJ, 251, 133
Moffat, A.F.J., Robert, C., 1992, in: "Nonisotropic and Variable Outflows from Stars", A.S.P. Conference Series, Vol.22, eds. L. Drissen, C. Leitherer & A. Nota, p.203
Owocki, S.P., 1994, ApSpSc, 221, 3
Puls, J., Kudritzki, R.-P., Herrero, A., Pauldrach, A.W.A., Haser, S.M., Lennon, D.J., Gabler, R., Voels, S.A., Vilchez, J.M., Wachter, S., Feldmeier, A., 1995, A&A, in press
Robert, C., 1992, Ph.D. Thesis, Département de Physique, Université de Montréal