1 Introduction

Cygnus X-3 has been known as a binary system since its discovery by Giacconi et al. (1967), but there is still debate about the masses of the two stars and the morphology of the system (for a review see Bonnet-Bidaud & Chardin 1988). The distance of the object is 8-12.5 kpc with an absorption on the line of sight $A_V\sim 20$ mag (van Kerkwijk et al. 1996). The flux modulation at a period of 4.8 hours, first discovered in X-rays (Parsignault 1972), then at near infrared wavelengths (Becklin et al. 1973), and observed simultaneously at X-ray and near-IR wavelengths by Mason et al. (1986), is believed to be the orbital period of the binary system. Following infrared spectroscopic measurements (van Kerkwijk et al. 1992), where WR-like features have been detected in I and K band spectra, the nature of the mass-donating star is suggested to be a Wolf-Rayet-like star, but an unambiguous classification, similar to the other WR stars, is still lacking. Mitra (1996) and Vanbeveren et al. (1998) pointed out that it is not possible to find a model that meets all the observed properties of Cygnus X-3 where the companion star is a normal Population I Wolf-Rayet star with a spherically symmetric stellar wind. In the evolution model originally proposed by van den Heuvel & de Loore (1973) a final period of the order of 4.8 h may result from a system with initial masses M₁⁰ = 15  ${M}_\odot$, M₂⁰ = 1  ${M}_\odot$, P⁰ = 5d, the final system being a neutron star accreting at a limited rate of $\sim$10^-7  ${M}_\odot$ yr^-1, from the wind of a core He burning star of about 3.8  ${M}_\odot$. Vanbeveren et al. (1998) proposed that the progenitor of Cygnus X-3 is a 50  ${M}_\odot$ + 10  ${M}_\odot$ system with P⁰ = 6 d; after spiral-in of the black hole into the envelope of the companion, the hydrogen reach layers are removed, and a 2-2.5  ${M}_\odot$ Wolf-Rayet like star remains with P = 0.2 d. A system containing a black hole and an He core burning star is also favored by Ergma & van den Heuvel (1998). In addition, Cygnus X-3 undergoes giant radio bursts and there is evidence of jet-like structures moving away from Cygnus X-3 at 0.3-0.9 c (Mioduszewski et al. 1998; 2001; Martí et al. 2001).
The main objective of the Infrared Space Observatory (ISO) spectrophotometric measurements in the 2.4-12 $\mu$m range was to constrain further the nature of the companion star to the compact object: the expected strong He lines as well as the metallic lines in different ionization states are important clues, together with the spectral shape of the continuum in a wavelength range as large as possible. An additional motivation for the imaging photometry with ISOCAM was to provide spatial resolution to a possible extended emission feature as a remnant of the expected high mass loss from the system. The paper is laid out as follows. In Sect. 2 observational aspects are reviewed. Section 3 summarizes the results on the continuum and line emissions from Cygnus X-3 and four Wolf-Rayet stars of WN 6, 7 and 8 types, and reviews the constraints set by the present observations on the wind and on the nature of the companion to the compact object in Cygnux X-3. Finally, Sect. 4 summarizes the conclusions of this paper.

 

Table 1: The ISO observing modes, ISO identifications, wavelength ranges and observing times on April 7, 1996, used on Cygnus X-3.