Quote from smAsh on January 4, 2020, 7:18 am

AAAnd, just read this one this morning.  I couldn't sleep, so I read about the tail of the swan.  Again.  This article is AWESOME!

Infrared spectroscopic variability of Cygnus X-3 in outburst and quiescence 

R. P. Fender, M. M. Hanson, G. G. Pooley

Monthly Notices of the Royal Astronomical Society, Volume 308, Issue 2, September 1999, Pages 473–484, https://doi.org/10.1046/j.1365-8711.1999.02726.x

Published:

19 September 1999

Abstract

We present four epochs of high-resolution infrared spectroscopy of the peculiar X-ray binary Cygnus X-3. The observations cover quiescent, small-flaring and outburst states of the system as defined by radio and X-ray monitoring. The underlying infrared spectrum of the source, as observed during radio and X-ray quiescence and small-flaring states, is one of broad, weak He II and N V emission. Spectral variability in this state is dominated by modulation at the 4.8-h orbital period of the system. H-band spectra confirm the significant hydrogen depletion of the mass donor. The closest spectral match to the quiescent infrared spectrum of Cyg X-3 is an early-type WN Wolf—Rayet star.In outburst, the infrared spectrum is dramatically different, with the appearance of very strong twin-peaked He I emission displaying both day-to-day variability and V(iolet)/R(ed) variations with orbital phase. We argue that the twin-peaked emission cannot arise in relativistic jets or, unless the distance to Cyg X-3 is severely overestimated, an accretion disc. The most likely explanation appears to be an enhanced stellar wind from the companion. Thus X-ray and radio outbursts in this system are likely to originate in mass-transfer, and not disc, instabilities, and the lengthening of the orbital period will not be smooth but will be accelerated during these outbursts. Furthermore, the appearance of these lines is suggestive of an asymmetric emitting region. We propose that the wind in Cyg X-3 is significantly flattened in the plane of the binary orbit. This may explain the observed twin-peaked He I features as well as reconciling a massive Wolf—Rayet secondary with the relatively small optical depth to X-rays, if the disc wind is inclined at some angle to the line of sight. A small set of observations following outburst, when the system was returning to a more quiescent X-ray and radio state, reveal strong He I 2.058-μm absorption with a clear P Cygni profile, at the same time as the more common weak He II and N V features. In a disc-wind geometry this can be interpreted as absorption in the densest, accelerating regions of the wind which can be viewed directly if the disc is inclined at some angle to the line of sight.

line: profiles, binaries: close, circumstellar matter, stars: individual: Cygnus X-3, infrared: stars

Issue Section:

Papers

1 Introduction

Cygnus X-3 is a heavily obscured luminous X-ray binary in the Galactic plane which displays a unique and poorly understood combination of observational properties. These include strong radio emission, with a flat spectrum extending to (at least) mm wavelengths in quiescence (e.g. Waltman et al. 1994; Fender et al. 1995) and giant flares which are associated with a relativistic jet (e.g. Geldzahler et al. 1983; Fender et al. 1997; Mioduszewski et al. 1998). In the infrared the system is bright with occasional rapid flare events and thermal continuum consistent with a strong stellar wind (e.g. Fender et al. 1996; van Kerkwijk et al. 1996). There is no optical counterpart at wavelengths shorter than ∼0.8 μm due to heavy interstellar extinction. The system is persistently bright in soft and hard X-rays (e.g. van der Klis 1993; Berger & van der Klis 1994; Matz et al. 1996), with strong and variable metal emission lines (e.g. Kawashima & Kitamoto 1996; Liedahl & Paerels 1996). Several detections at γ-ray energies have been claimed but rarely confirmed (see, e.g., Protheroe 1994). A clear and persistent (observed for >20 yr) asymmetric modulation in the X-ray and infrared continuum emission with a period of 4.8 h (e.g. Mason, Cordova & White 1986) is interpreted as the orbital period of the system. This period is rapidly lengthening with a characteristic time-scale of less than a million years (e.g. Kitamoto et al. 1995)

Infrared spectroscopy of the system in 1991 (van Kerkwijk et al. 1992) first revealed the presence of broad emission lines and an absence of hydrogen which was reminiscent of Wolf—Rayet stars. These observations have subsequently been confirmed and expanded upon (van Kerkwijk 1993; van Kerkwijk et al. 1996), and the binary interpreted as comprising a compact object (neutron star or black hole) and the helium core of a massive star, embedded within a dense stellar wind. Such an evolutionary end-point was predicted for Cyg X-3 as far back as 1973 by van den Heuvel & de Loore (1973). Unfortunately, most models of Wolf—Rayet stars do not envisage objects which can be contained within a 4.8-h orbit, causing some dispute over this interpretation (e.g. Schmutz 1993). Doppler-shifting of the broad emission lines with the orbital period of the system, with maximum blueshift at X-ray minimum, is interpreted by van Kerkwijk (1993) and van Kerkwijk et al. (1996) as being due to the lines arising in the region of the stellar wind shadowed from the X-rays of the compact object by the companion star. In this way the semi-amplitude of the Doppler shifts reflects only the wind velocity and gives no information on mass function of the system. Schmutz, Geballe & Schild (1996) interpret the Doppler-shifting of the emission lines with the orbital period more conventionally as tracking directly the motion of the companion star, and derive a mass function which implies the presence of a black hole of mass >10 M_⊙ in the system. However, their intepretation does not explain the phasing of the emission lines relative to the X-rays, nor is this discrepancy addressed in their work.

Mitra (1996, 1998) has argued that Cyg X-3 cannot contain a massive Wolf—Rayet star, as the optical depth to X-rays for a compact object in a tight 4.8-h orbit would be ≫1. The alternative explanation put forward is that Cyg X-3 instead contains a neutron star and an extremely low-mass dwarf, cf. PSR 1957+20.

Van Kerkwijk (1993) discussed the dramatic variability in line strengths and line ratios in the infrared spectra of Cyg X-3, and suggested that when the source is bright in X-rays the emission lines should be weak and orbitally modulated, but when the source is weak in X-rays the lines should be strong and show little orbital modulation. However, as noted in van Kerkwijk et al. (1996); Kitamoto et al. (1994) show that the strength of infrared line and X-ray emission are in fact probably broadly correlated from epoch to epoch, with the strong-lined spectrum of 1991 being obtained during an outburst of the system. The explanation put forward for this was enhanced mass-loss from the companion during outbursts, which both increases X-ray brightness (more accretion) and emission-line strengths. This model was combined with detailed radio, (sub)mm and infrared (photometric) observations obtained during an outburst, and expanded upon in Fender et al. (1997). Waltman et al. (1995) clearly indicate the epochs of the published infrared spectra against the Green Bank 2-GHz radio monitoring of the system.

In this paper we present four epochs of high-resolution infrared spectroscopy of Cyg X-3 with the Multiple Mirror Telescope over a two-year period. These observations cover periods of quiescence, small flaring and major outburst as revealed in radio and X-ray monitoring, and we discuss the clear changes in the spectrum of the source as a function of state. In a future paper we will analyse and discuss the results of our spectra that fully sample the entire orbit of Cyg X-3 during quiescence and during outburst.

2 Observations

2.1 Infrared

All observations were made using the Steward Observatory's infrared spectrometer, FSpec (Williams et al. 1993), on the Multiple Mirror Telescope (MMT). The spectra were taken using the medium-resolution, 300 groove mm⁻¹ grating, yielding a 2-pixel resolution element of 0.0018 μm, or R≈1200 at 2.12 μm, and R≈900 at 1.62 μm. The same observing procedure was used on all nights. A full log of these observations is provided in Appendix A.

The spectrometer has a slit size of 1.2×32 arcsec² on the MMT, allowing Cyg X-3 to be observed in four unique positions along the slit. In the reductions, after dark current had been subtracted and a flat-field divided from the raw two-dimensional images, sky emission and additional thermal background was removed by subtracting one slit position from the next. The integration times for Cyg X-3 were very long, either 2 or 4 min at each slit position (see tables in Appendix A). This is long enough that the strong atmospheric OH emission lines did not always subtract away cleanly due to temporal variations in atmospheric conditions between slit positions. In many cases, a few per cent scaling was required to get the OH features to disappear entirely. Background normalization of a few per cent was performed to remove fluctuations in thermal background between integrations.

Interspersed between our Cyg X-3 observations we obtained spectra of other stars, which were used to correct for telluric absorption features. The same telluric standard star, HR 7826, an A1V star, was used through out all observations. The intrinsic spectrum of the standard star, HR 7826, was determined using two secondary telluric standard stars, HR 7503 (16 Cyg A), a G1.5V star, and the O3 If star Cyg OB2 #7. A first estimate of the intrinsic spectrum of HR 7503 was obtained using a solar spectrum. The 2-μm spectrum of Cyg OB2 #7 is nearly featureless, with the exception of N III at 2.115 μm (Hanson, Conti & Rieke 1996). The A1V telluric standard contains only the Brγ feature. By ratioing these three spectroscopically unique telluric standard stars against each other, we were able to obtain a good determination of the intrinsic spectrum of each star. The intrinsic spectrum of HR 7826 was determined during our first observing run in 1996 June. This solution for the intrinsic spectrum was used throughout that run and will be used for all future observing runs. If our determination of the intrinsic spectrum of HR 7826 is not exactly correct, which is certainly the case at some level, any spurious features we have introduced will at least be consistently introduced into all of our Cyg X-3 spectra. This is important since it is our hope to study flux and velocity variations of very weak broad features in Cyg X-3 in a forthcoming paper.

Mean spectra for the four epochs of observation, and their relation to the changing X-ray and radio state of Cyg X-3, are shown in Fig. 1. Note that these spectra are not normalized, whereas those throughout the rest of the paper are. This is in order to show the approximate constancy of the continuum slope in different states. The correlation between radio flaring and bright X-ray states, originally proposed by Watanabe et al. (1994), is also obvious from Fig. 1.

 

Figure 1.

Open in new tabDownload slide

An illustration of our four observing periods, labelled A–D, against a backdrop of X-ray and radio monitoring. We have observed Cyg X-3 in distinct states of quiescence, small flaring and outburst. Spectra characteristic of each epoch are indicated in the lower panel, with tick marks indicating the lines identified in Fig. 2. The approximate S/N ratios for the spectra A–D are 80, 50, 100 and 40 respectively. Note that these spectra, unlike those presented in the rest of the paper, have not been normalized to the continuum; this is in order to show that there is no dramatic change in continuum slope from outburst to quiescence.

We began our first Cyg X-3 observing campaign in late May, 1996, which is symbolized in Fig. 1 as epoch A. For 10 of the 11 consecutive nights, Cyg X-3 was observed at approximately the same UT. Because a 24-h daily cycle is almost exactly five binary orbits, we were observing Cyg X-3 at close to the same orbital phase for these 10 nights (see Table 1 in Appendix A). Furthermore, on 1996 June 2, Cyg X-3 was observed over an entire orbital period, from φ_X=0.185 to 1.181 (quadratic ephemeris of Kitamoto et al. 1995, where φ_X=0 corresponds to minimum X-ray flux in the 4.8-h modulation, probably the point of superior conjunction of the compact object). During this first campaign, H-band spectra centred at 1.62 μm were also obtained on 1996 June 7 and 8 (see Fig. 2).

 

Table 1.

Open in new tabDownload slide

Line identifications and equivalent widths.

 

Figure 2.

Open in new tabDownload slide

Mean H-band (quiescent, taken 1996 June) and K-band (outburst, upper taken 1997 June; quiescence, lower taken 1996 June) spectra and line identifications.. UAF indicates a persistent (in 1996 June at least) unidentified absorption feature. The approximate S/N ratios for the H-band, K-band (quiescent) and K-band (outburst) spectra are 30, 100 and 80 respectively. These spectra, and those presented throughout the rest of the paper, have been normalized to the continuum.

The second campaign of observations, represented by epoch B in Fig. 1, began on 1996 September 22, where we obtained spectra covering the entire orbital period, from φ_X=0.382 to 1.429. One-fifth of an orbit was observed the following night (Table 2 in Appendix A). The third observing campaign covered five consecutive nights beginning on 1997 July 16 (Table 3 in Appendix A) and are represented in Fig. 1 as epoch C. The fourth night of the observations taken during epoch C covered one orbit, sampling from φ_X=0.205 to 1.109. Our final observing campaign, represented by epoch D in Fig. 1, covered just one-quarter of an orbital period on 1997 October 15.

 

Table A1.

Open in new tabDownload slide

1996 May/June observations.

 

Table A2.

Open in new tabDownload slide

1996 May/June observations.

2.2 Radio

The Ryle Telescope observations, at 15 GHz with a bandwith of 350 MHz, follow the pattern described in Fender et al. (1997). Data points shown in Figs 1 and 3 are 5-min integrations. The typical uncertainty in the flux-density scale from day to day is 3 per cent, and the rms noise on a single integration is less than 2 mJy.

 

Figure 3.

Open in new tabDownload slide

Spectra on five consecutive nights during outburst in 1997 June. Top panel shows XTE ASM 2–12 keV (open symbols) and 15-GHz radio monitoring (filled symbols), revealing the source to be in a bright and variable state, presumably arising from enhanced accretion and jet formation. All spectra are dominated by apparently twin-peaked He I emission. Day 1 corresponds to 1997 June 16. Days 4 and 5 reveal the development and subsequent decline of especially strong emission. This structure displays the same blue : red wing variability with orbital phase on two subsequent days, being twin-peaked around phase zero but very red-dominated half an orbit later.

2.3 Xte

Cyg X-3 is monitored up to several times daily in the 2–12 keV band by the Rossi XTE All-Sky Monitor (ASM). See, e.g., Levine et al. (1996) for more details. The total source intensity in the 2–12 keV band for individual scans is plotted in the top panels of Figs 1 and 3.

3 Line identifications

Line identifications in Cyg X-3 are shown in Fig. 2 and listed in Table 1. We display two different K-band spectra in Fig. 2, the upper taken during a time of high X-ray and radio activity, and the lower taken during quiescence. The strongest features include the 2.0587-μm He I singlet during outburst and the 2.1891-μm He II (77-4) during quiescence. The H-band spectrum centred at 1.62 μm, displays only a few identifiable features, He II (13−7) and (12−7) at 1.5719 and 1.6931 μm, respectively, and N V (10-9) at 1.554 μm. These H-band features were also evident in earlier UKIRT spectra from 1992 May 30, one day after K-band spectra revealed Cyg X-3 to be in a weak-lined state equivalent to quiescence as defined in this paper (M. H. van Kerkwijk, private communication). There is no evidence for any Brackett series hydrogen features. The H-band spectrum shown in Fig. 2 was taken in 1996 June, when Cyg X-3 was in a quiescent phase.

There is one absorption feature, centred at approximately 2.129 μm, that we have been unable to positively identify. It is unlikely that it is a feature due to intervening interstellar material, as numerous stars with line-of-sight extinction greater than 10 mag in the visible have been observed without ever showing such a feature (Tamblyn et al. 1996; Hanson, Howarth & Conti 1997; Watson & Hanson 1997). We suspect, then, that it must be related to the Cyg X-3 system. Curiously, it shows no shifting with the orbit, unlike the other lines in the K band (with the possible exception of He I at 2.058 μm). This unidentified absorption feature (UAF) has since disappeared from the spectrum, starting in 1997 June. We have seriously considered that the feature may be spurious, introduced by poor telluric corrections, or perhaps a bad pixel on the array. However, we see it present through out the entire 11-day run in 1996 June, despite small changes in grating position, against three different telluric standard stars, and new calibration images taken each day. Furthermore, inspection of earlier 2-μm spectra of Cyg X-3, while of lower resolution, seems to substantiate the presence of a weak absorption feature at 2.129 μm (van Kerkwijk et al. 1996). However, without an identification, we are unable to comment further on its nature or its possible relation to the Cyg X-3 system.

4 Spectral variability

In this section we discuss the observational properties at each of the four epochs for when near-infrared spectra were obtained. It is our aim to establish the spectral characteristics and nature of any variations seen in Cyg X-3 in different radio and X-ray states. This may help us to identify the origin of the spectral features, be they from the secondary star or the compact object. Spectra characteristic of each epoch are plotted in Fig. 1.

4.1 1996 May/June: quiesceuiescence

Represented by epoch A in Fig. 1, this is the longest continuous set of near-infrared observations ever taken of Cyg X-3. The source is in a state of radio and X-ray quiescence, with radio flux densities at 15 GHz in the range 40–140 mJy and XTE ASM fluxes in the range 4–9 count s⁻¹. The spectrum is dominated by broad, weak He II and N V emission, and weak, more narrow and intermittent He I (2.058 μm) absorption. Nearly all spectral variability is related to the 4.8-h orbital modulation, namely Doppler-shifting of the broad emission features. The full amplitude of the Doppler-shifting is of the same order as that reported by van Kerkwijk (1993) and Schmutz et al. (1996), i.e., 1000–1500 km s⁻¹. Orbitally phase-resolved spectra and dynamical interpretations will be presented elsewhere. The unidentified absorption feature (UAF) at 2.129 μm is also detected, but cannot be clearly identified with any known transition. The UAF shows no Doppler-shifting.

4.2 1996 September: small flaring

The second set of observation, epoch B in Fig. 1, caught Cyg X-3 in a more active phase. Radio observations at 15 GHz showed many small flares, with flux densities ranging from 50–450 mJy, corresponding to the ‘small-flaring’ state classified by Waltman et al. (1995). The XTE ASM recorded 8–17 count s⁻¹, significantly higher and more variable than in 1996 May/June. However, the K-band spectrum is very similar to that obtained at epoch A, showing little variability that is not orbitally related, and being dominated by the broad, weak He II and N V emission. The unidentified absorption feature at 2.129 μm appears to have weakened considerably in the three months since 1996 May/June.

4.3 1997 June: outburst

Epoch C represents observations during a major outburst of Cyg X-3. XTE ASM count rates varied rapidly between 14 and 32 count s⁻¹, having peaked at ≥40 count s⁻¹ around 100 days earlier. The radio emission was undergoing a second sequence of major flaring within 200 days. During the period of these observations flux densities of up to 3 Jy at 15 GHz were recorded. During the first period of radio flaring (MJD 50400–50500) Mioduszewski et al. (1998) clearly resolved an asymmetric, probably relativistic, jet from the source.

The K-band spectrum at this epoch is wildly different from that at any other epoch, being dominated by what appear to be very strong double-peaked He I emission features, most obviously at 2.058 μm. Significant day-to-day spectral changes which are not related to orbital phase are evident at this epoch, unlike in quiescence (where spectral variability is almost entirely due to orbital modulation — see above).

Fig. 3 illustrates the dramatic variability in the strength of the double-peaked He I emission over the five nights of observations: strongest emission is present on the fourth night, 1997 June 19. Fig. 4 shows in detail the rapid V/R variability, probably cyclic at the 4.8-h orbital period, observed on this date. Fig. 3 also hints at a possible anticorrelation between 2–12 keV X-ray flux and He I emission-line strength on day time-scales; while there is a large degree of X-ray variability in individual scans, the daily-averaged flux drops during the first four days to a minimum just after June 19. This is in contrast to the longer term correlation between X-ray state and He I line strength, and probably results from a drop in ionization state of the wind following a temporary decrease in X-ray flux.

 

Figure 4.

Open in new tabDownload slide

Rapid V/R variability in the He I 2p–2s (2.0857 μm) emission line on 1997 June 19 (day 4 in Fig. 3). The variability is probably cyclic at the 4.8-h orbital period. The red-shifted peak is far more persistent than the blueshifted peak, suggestive of some continuous moderately blueshifted absorption.

The dramatic day-to-day variability illustrated in Fig. 3 suggests that the lines seen are transient features which are not tied to a steady-state wind of the secondary. The wind is most likely changing in ionization state and/or density/velocity on day time-scales. The strongest peak seen in He I on day 4, showing up blueshifted at φ_X=0.0 and redshifted at φ_X=0.5, is clearly evolving on a time-scale of days, more than tripling in strength between days 3 and 4, and declining again within 24 h. The appearance of the 1.0830-μm He I feature on 1993 June 14 (van Kerkwijk et al. 1996) probably represented another one of these events, although 2-μm spectra are not available to confirm. Van Kerkwijk et al. (1996) do, however, show there was a marked increase in K-band flux on 1993 June 14. Such near-infrared flux increases on day time-scales have been seen during radio outbursts (Fender et al. 1997), and appear to be distinct from the more rapid (second to minute time-scales) infrared flaring which is often observed (e.g. Fender et al. 1996).

4.4 1997 October: post-outburst/small flaring

A single short (64 min total) observation on 1997 October 15 (φ_X=0.58–0.81), during an apparent decline to quiescence following outburst, epoch D, again reveals previously unobserved features. Alongside the quiescent weak broad He II and N V emission is strong He I 2.0587-μm absorption, displaying a P Cygni profile. This absorption is stronger than observed at any time during epoch A. The absorption is present in all individual spectra, and there is no evidence for significant variability on short (minutes) time-scales. The absorption minimum occurs within uncertainties at the rest wavelength of the transition, 2.058 μm, and the blue wing extends to ∼2.054 μm, implying a minimum outflow velocity of 500 km s⁻¹. The He I absorption feature does not display any Doppler-shifting, although our phase coverage is not ideal. There are no other significant absorption features in the spectrum. The UAF feature is entirely absent.

A comparison of the spectrum around the He I 2.0587 μm with that in outburst (Fig. 5) shows that the deep absorption may well be present in outburst also, but is completely dominated by much enhanced emission at this stage. This is compatible with the model for a disc-like wind, which we will explore in Section 6.

 

Figure 5.

Open in new tabDownload slide

A comparison of outburst (epoch C, 1997 June 19) and post-outburst (epoch D, 1997 Oct 15) spectra around the He I 2p–2s line at 2.0587 μm, summed in the phase interval 0.57–0.81. It seems plausible that the strong P Cygni absorption to −500 km s⁻¹, clearly evident in the post-outburst spectrum, is present at both epochs, and the only difference may be vastly reduced amount of emission in 1997 October.

5 Discussion

5.1 The near-infrared spectral type of the secondary

Van Kerkwijk et al. (1992)published the first near-infrared spectra of Cyg X-3, covering 0.72–1.05 μm (I band) and 2.0–2.4 μm (K band). These spectra, taken in late June 1991, displayed strong emission lines of He I and He II. The I-band spectrum in particular, showed a conspicuous absence of hydrogen lines. The lack (or much reduced fraction) of hydrogen, the strong He I emission at 2.058 μm, and the broad He II emission lines were interpreted as coming from the wind of the binary companion to the compact object in Cyg X-3. Based on the 1991 spectrum, and using comparison spectra obtained of several Wolf—Rayet stars which were observed at the same time, a spectral type of WN7 was estimated for the companion. There are some problems, however, with the 1991 June spectra. The spectrum showed strong narrow He I at 2.0578 μm with strong, broad He II, which is not generally seen in hydrogen-free WN Wolf—Rayet stars (Figer, McLean & Najarro 1997; cf. WR 123 in Crowther & Smith 1996). This subtle mis-match of spectral characteristics suggested that the lines seen in the original 1991 June spectrum did not originate solely from a WR-like wind. Indeed, subsequent spectra taken by van Kerkwijk (1993) showed that the originally strong He I features had since disappeared. These later spectra were now dominated by the broad He II features, as well as N V and N III. Such features are indicative of an earlier WR wind, perhaps WN4/5. However, as noted by van Kerkwijk et al. (1996), the line ratios between the nitrogen and He II lines are not consistent with such an early spectral class. In fact, the near-infrared He II lines in Cyg X-3 are extremely weak compared to other early WN stars (Crowther & Smith 1996; Figer et al. 1997).

We are now able to show that the original June 1991 spectrum of van Kerkwijk et al. (1992) was anomalous and almost certainly associated with an outburst in the system. Our 1997 June spectra are dominated by double-peaked emission, which does not seem to be traced in the original 1991 spectrum. However, by choosing a phase that was dominated by one peak and smoothing our spectra to the lower resolution of the van Kerkwijk et al. (1992) spectrum, our 1997 June spectra became a very close match in both lines detected and relative strength to the 1991 spectrum (Fig. 6). As already suspected by van Kerkwijk et al. (1996), the original K-band spectrum of van Kerkwijk et al. (1992) therefore appears to have been anomalous due to an outburst of Cyg X-3.

 

Figure 6.

Open in new tabDownload slide

A comparison of the 1991 June 29 spectrum of van Kerkwijk et al. (1992) with our spectrum of 1997 June 20, obtained during a radio and X-ray outburst. The similarity of the two spectra (contrast with the quiescent spectra in Figs 1 and 2) indicates that the initial WN7 classification based upon van Kerkwijk's spectrum was not representative of the underlying spectral type of the companion, but instead a result of enhanced He I emission during outburst.

The quiescent spectrum, dominated by weak, broad He II features, probably originates in the more steady-state wind of the stellar companion of Cyg X-3, and is our best diagnostic of the nature of this component. However, even this phase is not consistent with a normal WR wind. As first suggested by van Kerkwijk (1993), the presence of the high-energy compact object, circling the companion star at very close radii (estimated to be on the order of 5 R_⊙), has probably altered the wind structure of its companion (the ‘Hatchett McCray effect, ’Hatchett & McCray 1977). Stellar winds in early-type stars are driven through high-opacity resonance lines of such species as C IV and N V. The predominance of very high-energy photons from the compact object completely alters the ionization structure and thus the driving force of the wind, and may entirely eliminate significant line formation (McCray & Hatchett 1975). In the presence of the compact object, an X-ray-excited, thermally driven wind is instead created, which may have little or no line formation (Stevens 1991; Blondin 1994). Where the compact object is entirely blocked by the central disc of the companion, the expanding wind from the helium star may be capable of creating a normal line-driven wind, giving rise, at least weakly, to the broad high-ionization wind lines detected in Cyg X-3.

With such an interpretation for the line emission seen at near-infrared wavelengths, it would be difficult to infer many characteristics of the companion star. The most important characteristics of the companion star are that it is a helium-rich atmosphere, and it may be driving a fairly extensive, fast wind, both being reminiscent of late stages in massive star evolution. An early WN Wolf—Rayet star is probably the best candidate for the spectral type of the companion star. However, the mass of the companion star, and thus information on the mass of the compact object, can not be uniquely or confidently determined from the spectrum.

5.2 Twin-peaked He I emission in outburst

Here we discuss possible origins for the strong twin-peaked He I emission observed during outburst.

5.2.1 Jets?

Simple arguments show that the twin-peaked emission lines cannot arise directly from material in the relativistic outflows which are inferred from high-resolution radio mapping of Cyg X-3 (e.g. Geldzahler et al. 1983; Mioduszewski et al. 1998). First, the persistently stronger red wing as shown in Fig. 5 is the opposite of what would be predicted from Doppler-boosting, where the approaching (blueshifted) emission would be boosted, and the receding component diminished. Secondly, the relatively low velocity (∼1500 km s⁻¹) implied by the peak separation could only arise from a relativistic jet almost in the plane of the sky (i.e., with a very small radial component). In that case, the transverse Doppler shift due to time dilation would dominate, redshifting both components. For a velocity of around 0.3c this would result in a redshift of both components by ∼0.1 μm, which is not observed (note that this effect is observed in SS 433; see, e.g., Margon 1984). The lack of a discernible transverse redshift (assuming that the lines do correspond to the indicated He I transitions) effectively rules out a relativistic outflow.

A lower velocity, non-relativistic jet is possible, although it still suffers from the problem of explaining the persistently stronger red peak, but we consider this unlikely as rapidly variable radio emission is occurring throughout this period (Fig. 3). This is almost certainly associated with the production of a relativistic jet; given the existence of this jet and the strong wind, the presence of a third outflowing component (which matches the terminal velocity of the wind as inferred from quiescent observations) seems unlikely.

5.2.2 Accretion disc?

As already noted by van Kerkwijk et al. (1992), a significant contribution to the infrared emission of Cyg X-3 from an accretion disc is unlikely. This is because in order to generate the observed infrared luminosity the disc would need to be very hot (≥10⁶ K), as its size is tightly constrained by the 4.8-h orbit. Such a high temperature is hard to reconcile with the observed low-excitation He I features. However, the line profiles and velocity separation are reminiscent of features seen in optical spectra of accretion-disc-dominated systems, and it is worth checking in more detail.

We can calculate the temperature that a blackbody (the most efficient emitter) would require in order to reproduce the observed line flux, given that its size is constrained by the dimensions of the orbit. We assume a distance of 8.5 kpc, a binary separation of 5 R_⊙, a K-band extinction A_K=2.3 mag, and a flux in emission lines which is about 10 per cent of that in the continuum. For an observed flux density in the K band of ∼12 mJy (e.g. Fender et al. 1996) we find that we require a blackbody temperature in excess of 10⁶ K in order to produce the flux in the emission lines within the binary separation. As the emission of the plasma producing the lines is much less efficient than that of a blackbody, there seems to be no way in which the relatively low-excitation He I lines can be produced within the scale of the binary separation, as these lines need temperatures T≤10⁵ K (for reasonable densities). Using this temperature, we can find a minimum dimension for the emitting region. As r∝T^−1/2, we require an emitting region which is 3 times larger, i.e., ∼15 R_⊙=10¹² cm. Such a large separation for a 4.8-h orbit would imply a total mass in the system of 1000 M_⊙! So we can rule out an emitting zone which is contained within the orbit of the system.

Furthermore, the luminosity of the emission lines, both in outburst and quiescence, is orders of magnitude greater than that observed in K-band emission lines from the X-ray binary Sco X-1 (Bandyopadhyay et al. 1997). Given that Sco X-1, with a longer orbital period, probably possesses a larger (and hence brighter in the infrared) accretion disc than Cyg X-3, an origin for the infrared lines of Cyg X-3 in an accretion disc can be ruled out (unless the distance is overestimated by a factor of 10 or more — which seems highly unlikely, given the broad agreement between high optical/infrared extinction, high N_H in X-ray spectral fits, and the distance inferred from 21-cm radio observations).

So, in agreement with van Kerkwijk et al., we must conclude that the HeIemission lines arise from a region significantly larger than the binary separation of the system. This conclusion also rules out an origin for the emission lines in the X-ray-irradiated face of a relatively cool secondary.

5.2.3 An enhanced, possibly disc-like wind?

Here we discuss a third possible origin for the twin-peaked variable emission lines: a significant enhancement in the wind in Cyg X-3. This has already been suggested as the origin for outbursts from the system (Kitamoto et al. 1994; van Kerkwijk et al. 1996; Fender et al. 1997). Given that we have established that the twin-peaked emission lines almost certainly originate in an extended region which is not the jets, and the existing evidence for a strong wind in the Cyg X-3 system, a natural explanation is that the increased line strength in outburst represents an increase in the density of the WR-like wind in the system. Such an increase in density will be coupled to a decrease in the mean ionization level of the wind, hence the much increased He I : He II ratio. While an enhanced wind density of the companion star is a natural explanation for bright X-ray/radio states which reflect increased rates of accretion and jet formation, such enhancements have never been observed in other WN stars. Cyg X-3, however, is an exceptional system. It experiences both extreme tidal forces and irradiation, which probably induce erratic behaviour and non-periodic variations in the extended atmosphere of the companion helium star.

The appearance of the twin-peaked lines, and their variability (probably) in phase with the 4.8-h orbit suggests an origin in an asymmetric emitting region. We believe that this wind may be flattened and disc-like, probably in the plane of the binary (see, e.g., Stee & de Araújo 1994 for predicted line profiles from a disc-wind). A flattened wind may have formed in the Cyg X-3 system as a result of a rapidly (synchronously) rotating mass donor and/or focusing of non-accreted material into the binary plane by the compact object. In this case most of the infrared emission arises from material in the plane of the binary but outside the orbit, and the optical depth along the line of sight from the X-ray source to the observer remains small as long as the system is not viewed edge-on (see Fig. 7). In this way, the problem of reconciling the Wolf—Rayet spectral typing of the companion with the detection of X-ray emission from near the centre of the system, highlighted by Mitra (1996; 1998), can be side-stepped whilst also explaining the high infrared luminosity of the system. It is worth recalling, however, that Berger & van der Klis (1994) show from timing studies that the X-ray emission from Cyg X-3 must still be undergoing significant scattering.

 

Figure 7.

Open in new tabDownload slide

A disc-wind in the Cyg X-3 system. In quiescence orbital modulation follows essentially the model of van Kerkwijk (1993) and van Kerkwijk et al. (1996), with the X-ray source ionizing the entire wind except that region shadowed by the companion. In outburst, caused by enhanced mass-loss from the companion star, the X-ray source can only ionize a small local region (Strömgren zone) and He I emission dominates. V/R variability and asymmetry is caused by P Cygni absorption from the accelerating region of the wind seen against the companion, and by the Strömgren zone as it tracks the X-ray source around the 4.8-h orbit. This disc-wind model can reconcile a massive, Wolf—Rayet-like companion with a small optical depth to X-rays.

Further support for a disc-wind model may come from the infrared polarimetric observations of Jones et al. (1994), who found a significant degree of intrinsic polarization from Cyg X-3 in the infrared K band. They suggested that this may indicate a preferential plane of scattering in the binary. Several WR stars also show intrinsic polarization, interpreted as arising from scattering in a flattened wind (e.g. Schulte-Ladbeck, Meade & Hillier 1992; Schulte-Ladbeck 1995; Harries, Hillier & Howarth 1998). Such intrinsic polarization seems to be more common from WN-type Wolf—Rayets (Schulte-Ladbeck et al. 1992; Harries et al. 1998); and the only direct observation (radio interferometry) of a flattened Wolf—Rayet wind was also from a WN subtype (Williams et al. 1997). Additionally, the position angle of the radio jet in Cyg X-3 (e.g. Mioduszewski et al. 1998) is approximately perpendicular to the long axis of the flattened wind as inferred from the position angle (0°–40°) of the derived intrinsic infrared polarization. Assuming that the jet propagates along the axis of the accretion disc, which itself lies in the binary plane, this supports a model in which binary and wind planes are aligned.

We discuss the interpretation of the outburst state in the context of a flattened disc-like wind in more detail below.

6 Orbital modulations with a disc-wind

6.1 Quiescence

Our disc-wind model for Cyg X-3 is sketched in Fig. 7. In such a model, Doppler-shifting of He II and N V lines in quiescence would occur essentially as outlined in the model of van Kerkwijk (1993) and van Kerkwijk et al. (1996) (hereafter the ‘van Kerkwijk’ model). We expect most of this emission to arise from outside the binary, with the compact object orbiting within the wind-accelerating zone of the WR-like companion. As discussed in the introduction, and in van Kerkwijk et al. (1996) and Mitra (1998), the van Kerkwijk model naturally explains the phasing of the X-ray and infrared continuum modulation with the Doppler shifts, in contrast to the model of Schmutz et al. (1996).

6.2 Outburst

During outburst, we presume that the much-enhanced mass-loss and consequent higher wind density prevents the X-ray source from ionizing anything but a small fraction of the wind (unlike in the van Kerkwijk model for quiescence in which the majority of the wind is ionized). A quantitative level of enhancement above the quiescent state is difficult to estimate. A realistic model to describe the increased He I emission would require knowledge of the geometry and structure (the clumpiness) of this wind, as well as the fractional increase in mass-loss rate. An increase in the soft X-ray flux by a factor of 3 during outburst indicates a corresponding increase in the mass accretion rate during such periods, although density enhancements close to the compact object may not exactly reflect those in the wind as a whole. It might be possible, given arguments based on the time-scale of the line structures seen, to estimate a fractional density of the wind during outburst. Such an in-depth analysis is beyond the scope of this study. Possibly the most accurate measure of the degree of wind enhancement may come from measurements of ‘glitches’ in the orbital period derivative, as angular momentum is lost from the system at a higher rate during the outbursts.

The rapid, probably cyclic V/R variability observed during outburst could occur as a combination of three components:

1. broad He I emission from the entire disc-wind, from approximately −1500 to +1500 km s⁻¹;
2. an ionized region (Strömgren zone) local to the X-ray source which depletes the He I emission in that region of the orbit, and
3. lower velocity (∼ −500 km s⁻¹) blueshifted (P Cygni) absorption from the accelerating region of the wind observed against the companion star.

This simple scheme (illustrated in the lower panel of Fig. 7) can qualitatively explain the observed phasing of the V/R variability and the greater persistence of the red-shifted peak:

Phase 0.0: X-ray source is on far side of wind from the observer. The red-shifted peak is depleted at relatively low velocities due to ionized zone around X-ray source. Similarly, the blueshifted peak is depleted at lower velocities due to persistent P Cygni absorption.

Phase 0.5: X-ray source is on near side of wind. Blue-shifted emission is depleted both by P Cygni absorption and ionization from X-ray source; the red-shifted peak is unaffected by either and is much stronger.

In the context of this model, the spectrum obtained in 1997 October (epoch D, Fig. 5) still shows deep P Cygni absorption but much-reduced emission. This may represent an intermediate state in the return to quiescence in which the low-velocity absorption is still occurring in the densest parts of the wind, but beyond the binary orbit most of the material is ionized and He II/N V dominate over He I as in quiescence.

7 Conclusions

We have presented the most comprehensive and highest resolution set of infrared spectra of Cyg X-3 to date. In combination with X-ray and radio monitoring, we can characterize the infrared spectral behaviour of the source in outburst and quiescence.

The underlying infrared spectrum of Cyg X-3, observed during both radio and X-ray outburst and quiescence, displays weak, broad, He II and N V (but no He I) emission. Some He I 2.058-μm absorption may be present, preferentially around orbital phase zero. H-band spectra extend our spectral coverage and confirm the significant He-enrichment of the mass donor, with no evidence of any hydrogen features. While not perfect, the closest match to the spectrum is that of a hydrogen-depleted, early WN-type Wolf—Rayet star.

In outburst, the K-band spectrum becomes dominated by twin-peaked He I emission, which is shown to be unlikely to arise in relativistic jets or an accretion disc. This emission seems to arise in an enhanced wind density, presumably also responsible for the X-ray and radio outburst via enhanced accretion and related jet formation. This explains the observed long-term (outburst time-scale) correlation between emission-line strength and X-ray and radio state, as noted in Kitamoto et al. (1994). The emitting region almost certainly extends beyond the binary orbit, and displays significant day-to-day intensity variations, as well as V/R variability with orbital phase. The short-term (day-to-day) variability in He I line strength may be anticorrelated with X-ray flux due to a varying degree of ionization of the wind. It seems that, for Cyg X-3 at least, the major X-ray and radio outbursts are due to mass-transfer, and not disc, instabilities. If this interpretation is correct, then the period evolution of Cyg X-3, determined by extreme mass-loss from the system (van Kerkwijk et al. 1992; Kitamoto et al. 1995) will not be smooth, instead displaying periods of accelerated lengthening during outbursts. The detection and measurement of such ‘glitches’ would be important both for understanding the evolution of the Cyg X-3 system and for estimating the amount of additional circumstellar material present during outbursts.

The appearance and variability of the emission features in outburst is suggestive of an asymmetric emitting region, and we propose that the wind in Cyg X-3 is significantly flattened, probably in the plane of the binary orbit. This may explain the intrinsic polarization of the infrared emission from Cyg X-3, which indicates a scattering plane perpendicular to the radio jet axis. The interpretation of a flattened wind is supported by polarimetric and direct radio interferometric observations revealing evidence for flattened winds in other Wolf—Rayet stars. A simple model for the V/R variability in outburst, in the context of a flattened disc-wind, comprising a small ionized zone around the compact object and continuous P Cygni absorption which erodes the blueshifted wing, qualitatively explains the observations. Furthermore, a disc-like wind in the Cyg X-3 system also naturally explains why we can have both a high infrared luminosity and yet still observe the X-ray source, a problem highlighted by Mitra (1996; 1998) as being very serious for a spherically symmetric wind. While there is still significant scattering of the X-rays along the line of sight (see Berger & van der Klis 1994) it will be considerably less than in the case of a spherically symmetric wind. Additionally, we note that the apparent one-sidedness of the radio jet from Cyg X-3 in the latest VLBA observations (Mioduszewski et al. 1998) may arise not from a jet aligned near to the line of sight (implying a nearly face-on orbit which is seemingly incompatible with the strong orbital modulations observed) but instead from the obscuration of the receding (northerly) jet by the far side of the disc-wind. This would naturally explain why the jet is so apparently one-sided on small scales and yet symmetrical on larger scales.

To conclude, the combination of a WR-like spectrum, high luminosity (M_K≤−5) and evidence for a disc-like wind supports the interpretation of Cyg X-3 as a high-mass X-ray binary in a very transient phase of its evolution.

Acknowledgments

We thank George and Marcia Rieke for the use of their near-infrared spectrometer and help during the observations. RPF thanks Rudy Wijnands for help with the XTE ASM light curves, and Marten van Kerkwijk, Elizabeth Waltman, Michiel van der Klis, Simon Clark, Jan van Paradijs, Lex Kaper and Rens Waters for many useful discussions. The MMT is jointly operated by the Smithsonian Astrophysics Observatory and the University of Arizona. We thank the staff at MRAO for maintenance and operation of the Ryle Telescope, which is supported by the PPARC. RPF was supported during the period of this research initially by ASTRON grant 781-76-017, and subsequently by EC Marie Curie Fellowship ERBFMBICT 972436. MMH has been supported by NASA through Hubble Fellowship grant #HF-1072.01-94A awarded by the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., for NASA under contract NAS 5-26555.

References

Bandyopadhyay

R.

Shahbaz

T.

Charles

P. A.

Van Kerkwijk

M. H.

Naylor

T.

,

1997

,

MNRAS

,

285

,

718

Crossref

Search ADS

Berger

M.

Van Der Klis

M.

,

1994

,

A&A

,

292

,

175

Blondin

J. M.

,

1994

,

ApJ

,

435

,

756

Crossref

Search ADS

Crowther

P.

Smith

L.

,

1996

,

A&A

,

305

,

541

Fender

R. P.

Bell Burnell

S. J.

Garrington

S. T.

Spencer

R. E.

Pooley

G. G.

,

1995

,

MNRAS

,

274

,

633

Crossref

Search ADS

Fender

R. P.

Bell Burnell

S. J.

Williams

P. M.

Webster

A. S.

,

1996

,

MNRAS

,

283

,

798

Crossref

Search ADS

Fender

R. P.

Bell Burnell

S. J.

Waltman

E. B.

Pooley

G. G.

Ghigo

F. D.

Foster

R. S.

,

1997

,

MNRAS

,

288

,

849

Crossref

Search ADS

Figer

D. F.

McClean

I. S.

Najarro

F.

,

1997

,

ApJ

,

486

,

420

Crossref

Search ADS

Geldzahler

B. J.

et al.

,

1983

,

ApJ

,

273

,

L65

Crossref

Search ADS

Hanson

M. M.

Conti

P. S.

Rieke

M. J.

,

1996

,

ApJS

,

107

,

281

Crossref

Search ADS

Hanson

M. M.

Howarth

I. D.

Conti

P. S.

,

1997

,

ApJ

,

489

,

698

Crossref

Search ADS

Harries

T. J.

Hiller

D. J.

Howarth

I. D.

,

1998

,

MNRAS

,

296

,

1072

Crossref

Search ADS

Hatchett

S.

McCray

R.

,

1977

,

ApJ

,

211

,

552

Crossref

Search ADS

Jones

T. J.

Gehrz

R. D.

Kobulnicky

H. A.

Molnar

L. A.

Howard

E. M.

,

1994

,

AJ

,

108

,

605

Crossref

Search ADS

Kawashima

K.

Kitamoto

S.

,

1996

,

PASJ

,

48

,

L113

Kitamoto

S.

Miyamoto

S.

Waltman

E. B.

Fiedler

R. L.

Johnston

K.

Ghigo

F. D.

,

1994

,

A&A

,

281

,

L85

Kitamoto

S.

et al.

,

1995

,

PASJ

,

47

,

233

Levine

A. M.

Bradt

H.

Cui

W.

Jernigan

J. G.

Morgan

E. H.

Remillard

R. A.

Shirey

R.

Smith

D.

,

1996

,

ApJ

,

469

,

L33

Crossref

Search ADS

Liedahl

D. A.

Paerels

F.

,

1996

,

ApJ

,

468

,

L33

Crossref

Search ADS

Margon

B.

,

1984

,

ARA&A

,

22

,

507

Crossref

Search ADS

Mason

K. O.

Cordova

F. A.

White

N. E.

,

1986

,

ApJ

,

309

,

700

Crossref

Search ADS

Matz

S. M.

Fender

R. P.

Bell Burnell

S. J.

Groove

J. E.

Strickman

M. S.

,

1996

,

A&AS

,

120C

,

235

McCray

R.

Hatchett

S.

,

1975

,

ApJ

,

199

,

196

Crossref

Search ADS

Mioduszewski

A. J.

Hjellming

R. M.

Rupen

M.

Waltman

E. B.

Pooley

G. G.

Ghigo

F. D.

Fender

R. P.

,

1998

, in

Zensus

J. A.

Taylor

G. B.

Wrobel

J. M.

, eds, ASP Conf. Ser. Vol.

144

,

Radio Emission from Galactic and Extragalactic Compact Sources

.

Astron. Soc. Pac.

,

San Francisco

, p.

351

Mitra

A.

,

1996

,

MNRAS

,

280

,

953

Crossref

Search ADS

Mitra

A.

,

1998

,

ApJ

,

499

,

385

Crossref

Search ADS

Protheroe

R. J.

,

1994

,

ApJS

,

90

,

883

Crossref

Search ADS

Schmutz

W.

,

1993

, in

Regev

O.

Shaviv

G.

, eds,

Ann. Israel Phys. Soc. 10, Cataclysmic Variables and Related Physics

.

IOP Publishing

,

Bristol

, p.

259

Schmutz

W.

Geballe

T. R.

Schild

H.

,

1996

,

A&A

,

311

,

L25

Schulte-Ladbeck

R.

,

1995

, in

Van Der Hucht

K. A.

Williams

P. M.

, eds, Proc. IAU Symp. 163,

Wolf—Rayet Stars: Binaries, Colliding winds, Evolution

.

Kluwer

,

Dordrecht

, p.

176

Schulte-Ladbeck

R.

Meade

M. R.

Hiller

D. J.

,

1992

, in

Drissen

L.

Leitherer

C.

Nota

A.

, eds, ASP Conf. Ser. Vol.

22

,

Non-isotropic and Variable Outflows from Stars

.

Astron. Soc. Pac.

,

San Francisco

, p.

118

Stee

PH.

De Araújo

F. X.

,

1994

,

A&A

,

292

,

221

Stevens

I. R.

,

1991

,

ApJ

,

379

,

310

Crossref

Search ADS

Tamblyn

P.

Rieke

G.

Hanson

M. M.

Close

L. M.

McCarthy

D. W.

Rieke

M. J.

,

1996

,

ApJ

,

456

,

206

Crossref

Search ADS

Van Den Heuvel

E. P. J.

De Loore

C.

,

1973

,

A&A

,

25

,

387

Van Der Klis

M.

,

1993

,

Space Sci. Rev.

,

62

,

173

Crossref

Search ADS

Van Kerkwijk

M. H.

,

1993

,

A&A

,

276

,

L9

Van Kerkwijk

M. H.

et al.

,

1992

,

Nat

,

355

,

703

Crossref

Search ADS

Van Kerkwijk

M. H.

Geballe

T. R.

King

D. L.

Van Der Klis

M.

Van Paradijs

J.

,

1996

,

A&A

,

314

,

521

Waltman

E. B.

Fiedler

R. L.

Johnston

K. J.

Ghigo

F. D.

,

1994

,

AJ

,

108

,

179

Crossref

Search ADS

Waltman

E. B.

Ghigo

F. D.

Johnston

K. J.

Foster

R. S.

Fiedler

R. L.

Spencer

J. H.

,

1995

,

AJ

,

110

,

290

Crossref

Search ADS

Waltman

E. B.

Foster

R. S.

Pooley

G. G.

Ghigo

F. D.

,

1997

,

Vistas Astron.

,

41

,

43

Crossref

Search ADS

Watanabe

H.

Kitamoto

S.

Miyamoto

S.

Fiedler

R. L.

Waltman

E. B.

Johnston

K. J.

Ghigo

F.

,

1994

,

ApJ

,

433

,

350

Crossref

Search ADS

Watson

A.

Hanson

M. M.

,

1997

,

ApJ

,

490

,

L165

Crossref

Search ADS

Williams

D.

Thompson

C. L.

Rieke

G. H.

Montgomery

E.

,

1993

,

Proc. SPIE

,

1308

,

482

Williams

P. M.

Dougherty

S. M.

Davis

R. J.

Van Der Hucht

K. A.

Bode

M. F.

Setia Gunawan

D. Y. A.

,

1997

,

MNRAS

,

289

,

10

Crossref

Search ADS

Appendix

Appendix A: Observing logs

Tables A1–4 below list the epochs and exposure times of every K- and H-band spectrum taken during all four observing runs. The reduced spectra are downloadable from

 

Table A3.

Open in new tabDownload slide

1997 June observations.

 

Table A4.

Open in new tabDownload slide

1997 October observations.

ftp://cdsarc.u-strasbg.fr/pub/cats/J/MNRAS/308/473

HJD, as used below, is heliocentric-corrected Julian Date, with 245 0000.0 subtracted. The ‘Exposures’ column shows three numbers, indicating the number of spectra averaged, the number of exposures and the length of each exposure (in seconds). Orbital phase at the start of each observation is indicated by φ.

We request that any use of these spectra in future publications makes reference to this work.

Quote from smAsh on January 4, 2020, 7:20 am

An article from December, 2019.

 

Recent Swift Spectroscopy of Cygnus X-3 ATel #13342; Y. Zhang, B. Tetarenko, J. M. Miller (Univ. of Michigan) on 10 Dec 2019; 15:10 UT Credential Certification: Jon Miller (jonmm@umich.edu) Subjects: X-ray, Binary, Black Hole, Transient Cygnus X-3 is a high mass X-ray binary with a compact object and a Wolf-Rayet companion. It has often been monitored using the Neil Gehrels Swift Observatory, and a new observation was obtained on 2019-11-15. We reduced the XRT data using the standard HEASOFT tools. The filtered "windowed timing" mode data have a count rate of 23.66 c/s. After binning to require a minimum of 15 counts per bin, we fit the data over the 1-10 keV band using XSPEC. Fitting with a model consisting of tbabs*(gauss+gauss+gauss+diskbb), with the Gaussian lines fixed at neutral, He-like, and H-like values (6.40 keV, 6.70 keV, and 6.97 keV), a chi-squared of 1.28 was found. The blackbody was measured to be extremely hot (kT = 3.7 keV), and the emitting area to be extremely small; this indicates that the component was trying to mimic a power-law or Comptonization component. There is also a small offset in each of the lines relative to the model; this likely indicates an instrumental shift. Finally, there are negative residuals in the 2-5 keV band, potentially indicating partial covering absorption. We next modeled the spectrum with tbabs(zmshift*(gaussian + gaussian + gaussian) + tbpcf*powerlaw). Here, the "zmshift" component simply accounts for a shift in the energy scale of the instrument; with this change, the iron lines are fit quite well. This model has a reduced chi-squared of 1.26, and it is more physically plausible and self-consistent than the previous model. The observed flux is measured to be 2.8 E-9 erg/cm^2/s; after removing the internal and line-of-sight obscuration, an unabsorbed flux of 8.4 E-9 erg/cm^2/s is implied, translating to a luminosity of 5.6 E+37 erg/s (for a distance of 7.4 kpc, see McCollough et al. 2016). This model may be a plausible means of characterizing many of the archival Swift monitoring observations of Cygnus X-3, and examining the flux of the Fe emission lines versus several parameters of interest. We will look to systematically apply it in a future publication. McCollough, M., et al., 2016, ApJ, 830, L36

---

---

[ Telegram Index ]

R. E. Rutledge, Editor-in-Chief
Derek Fox, Editor
Mansi M. Kasliwal, Co-Editor

Quote from smAsh on January 4, 2020, 7:23 am

Cygnus as related to ancient civilizations/ Gobekli Tepe

by Andrew Collins

2006

from AndrewCollins Website

 

 

A suspected Black Hole in Cygnus Constellation

Thought to be source of cosmic rays  that changed evolution and kick-started religion

excerpts from An Interview with Andrew Collins by Brent Raynes

Editor: Let's talk a little more about your upcoming book The Cygnus Mystery.

Andrew Collins: Right.

 

My current project is The Cygnus Mystery. Cosmic rays from the Cygnus constellation influenced human evolution in early history and affected human behavior and caused germ-like mutations over a longer period of time, probably about 2000 years.

 

Now this might seem a whacky idea but as I was writing the book I discovered that a scientific think tank, known as the Meinel Institute of Las Vegas announced that they also now believe that something similar was going on in paleolithic times.

 

Their candidate is a planetary nebula in the constellation of Draco, which is next door to Cygnus. Very, very close.

 

This they worked out through an examination of the levels of Beryllium-10, which is a bi-product of cosmic rays hitting the upper atmosphere and changing oxygen and nitrogen into the secondary element of beryllium. Which falls to the earth and is retained within sea levels and ice cores, and measurements of this can determine the level of cosmic rays.

 

There's no question that towards the end of the paleolithic era the level of cosmic rays hitting the earth had doubled. So to suggest that this may have been involved in mutations leading in changes to human evolution is not a mad idea.

 

Plus I found out that as early as 1972, Carl Sagan, the great scientific writer, spoke about exactly the same thing, the idea of cosmic rays influencing human evolution, in his book The Cosmic Connection. He went back to the subject in 1977 in his book The Dragons of Eden.

It's something that I believe influenced the foundations of world religions as well. You will find this Cygnus link within Christianity, within Islam, within Hinduism, within Near Eastern religions like Mandaeans of Iraq and Iran, the Sabians, the Chaldeans of the Bible, even certain Shiite Islamic groups strongly related to this.

 

It's behind the whole concept that heaven was in the north and that we come from the sky and that we return there in death.

 

Somehow this is all mixed up with this whole idea of cosmic rays influencing human evolution, influencing human intelligence, but mixed up with the whole idea of panspermia as well.

 

This connection is with the cosmos. Something which is very, very important, and I think that as we go on we will begin to realize that it is more and more important.

 

from 365Articles Website

 

 

Ever since the discovery in the 1920s that radiation can cause gene mutations, scientists have speculated on the role that high energy cosmic rays might have played in evolution.

 

Indeed, as early as 1930 it became the theme of a science-fiction story in which cosmic rays were harnessed by a mad scientist in order to rapidly transform himself into a super being millions of years ahead of his time,

While similar ideas must have been behind the entrance of the alien black monolith among a community of ape-men in Arthur C. Clarke's classic '2001: A Space Odyssey'.

 

Moreover, one of the greatest scientific minds of the twentieth century, American astronomer Carl Sagan wrote in 1973 that human evolution was the result of incoming cosmic rays from some distant neutron star, demonstrating how everything in the universe affects everything else.

 

It was a bold notion, but one destined not to find favour among geneticists, simply because there was no hard evidence that cosmic rays - first confirmed during a series of balloon ascents in 1912 by Austrian physicist Victor F. Hess (1883-1964) - have any real impact on evolution, whatever their origin (since there is no consensus on this fact).

Indeed, H.J. Muller (1890-1967), the American geneticist whose work with the fruit fly Drosophila led to the realization that radiation (he used X-rays and later radium) was a mutagen, addressed the topic in a paper published in 1930 and again in 1952.

 

He concluded that the cosmic ray flux penetrating the upper atmosphere and reaching ground level was inadequate to explain spontaneous mutations in life forms, whatever their type.

 

Muller was not wrong, but had he been privy to modern scientific data that clearly demonstrates that at certain times in the Earth's history it has been bombarded with high levels of cosmic rays then he might have thought again.

 

 

Records from the Ice

Information of this order comes from the fact that when so-called 'primary' cosmic rays hit the upper atmosphere almost all of them break up when they collide with nuclei of oxygen and nitrogen, the process producing a plethora of charged secondary particles.

 

Many disintegrate in milliseconds, but others with much longer half lives plummet to the ground and are preserved in everything from lake sediments to stalagmites and, more crucially, the ice that forms to great depths in the Arctic and Antarctic regions. One such isotope called Beryllium 10 (10Be) is clearly traceable in ice cores.

 

Since individual layers of ice form each year the levels of Beryllium 10 can be counted to provide accurate indications of cosmic ray activity in the upper atmosphere.

In recent years, an analysis of ice cores extracted from polar stations in Greenland and Antarctica have clearly demonstrated that over the past 100,000 years, there have been three periods when the cosmic ray flux has increased dramatically.

The first was around 60,000 years ago, the second occurred approximately 40,000-35,000 years, and the third and last peak began around 17,000 years ago. Each one lasted for a period of approximately 2,000 years. Similar results have been determined from a stalagmite removed from a submerged blue hole in the Bahamas.

 

An examination of its Beryllium-10 content indicates that between 45,000 and 11,000 years ago the Earth was bombarded by twice the amount of cosmic radiation than it is today.

 

Where's the Cosmic Source?

The first question one must ask is where this influx of cosmic radiation might have come from.

 

Was it really a neutron star, as Carl Sagan suggested, or could it have been another astronomical source such as a black hole, which produces jets of particles that reach out far across galaxies?

Alternatively, was there some other, more prosaic solution to this enigma? The more of less regular gaps between the spikes of Beryllium 10 activity noted in the ice cores might well indicate some kind of cyclic force in action, most obviously that of the sun.

 

Cosmic rays are known to be partially deflected by the solar magnetic field that stretches out far into the heart of the solar system, and it is believed that the rate of Beryllium 10 production in the upper atmosphere is dependent on the strength of the solar field, which is itself connected with sunspot activity.

 

Was it an Exploding Star?

In addition to this, the sun's long term climate cycles of 100,000, 41,000 and 23,000 years, first noted by Serbian geophysicist Milutin Milankovic (1879-1958), must also affect the production of Beryllium-10 for similar reasons, i.e. the influence of the solar field upon the Earth's upper atmosphere.

 

This said, there might easily have been other factors behind the sudden increase in cosmic rays hitting the earth, the most catastrophic being a supernova, the death of a star as it expels the last of its nuclear fuel and collapses to form a high-mass compact object, either a white dwarf, black hole or neutron star.

 

Supernova remnant SN1987A, which
exploded in 1987 and is here seen by the Hubble telescope in 1995.

(Credit: NASA/Hubble).

 

Supernovas produce enormous bursts of gamma rays and cosmic rays which are sent out across dozens if not hundreds of light years of space.

 

If such an event occurred close enough to our own solar system then the Earth would be showered by deadly radiation, which might damage the ozone layer, causing not only many more rays to reach ground level, but also the onset of high levels of UV radiation from the sun.

 

More conservatively, catastrophists suggest that a close supernova would send a barrage of cosmic particles in our direction, which would dramatically increase cloud formation, preventing the sun from penetrating through the atmosphere, and thus bringing about a sudden ice age.

Whatever the consequences of a close supernova, life on Earth would suffer mass extinctions. As terrifying a scenario as this might seem, it was the favoured theory for the sudden disappearance of the dinosaurs some 65 million years ago until the discovery in 1980 of the Chicxulub impact crater in Mexico's Yucatan peninsular.

This helped confirm the alternative theory that a super-sized asteroid or comet had been responsible for their extinction. Indeed, this was the opinion of Carl Sagan and his co-author Dr I.S. Shklovskii, the famous Soviet astrophysicist and radio astronomer, in a scholarly book entitled Intelligence in the Universe, published in 1966.

 

In fact, one wonders whether Sagan's unique view that cosmic rays accelerate human evolution actually stemmed from his obvious fascination with the extinction of the dinosaurs, even though we now know they were not killed off this way.

Yet the powerful idea of a close supernova wreaking devastation on earth during some past geological age lingers, with some catastrophists believing that it could have brought about mass extinctions during other geological epochs, for instance at the close of the Jurassic age some 145 million years ago, as well as at the culmination of the Pleistocene age, which coincided with the end of the last Ice Age, some 11,000 years ago.

 

And such scientific speculation is where it starts getting interesting, for when the high levels of beryllium 10 were first noted in the ice cores at the beginning of the 1990s, scientists from the Cosmic Ray Council of the Soviet Academy of Sciences, working alongside a team from the University of Arizona, speculated that they resulted from a supernova explosion just 150 lights years away.

 

That's just 900 million, million miles from here...

 

Cygnus Loop or Veil

the remnant of a supernova
(Pic credit: NASA).

To back up their dramatic claims the joint Soviet-American team cited the presence at around 150 light years away in the northern constellation of Cygnus of an immense formation of glowing clouds of gaseous debris - the remnants of an immense supernova explosion - known to astronomers as the Cygnus Veil, or Veil nebula.

Was this the remnants of the supernova explosion that had showered the Earth with cosmic rays for up to 2,000 years some 40,000-35,000 years ago? Did it bring about dramatic climatic changes and bursts of radiation that evolved humanity into what we are today?

The Emergence of Man

For whatever reason, the worldwide press coverage that resulted from this dramatic announcement of a close supernova some 35,000 years ago came to nothing.

 

Yet, thankfully, there was one person who did take notice, and this was British anthropological writer Denis Montgomery.

 

Having lived in Africa for many years, where anatomically modern humans emerged for the first time some 200,000 years ago, he became intrigued as to why sudden jumps of evolution occur.

 

Was it purely spontaneous, through chemical changes in the body, or were there other exterior factors at play, such as environmental and climatic changes, nutritional variety or even simple competitiveness?

Even though there is ample evidence that our earliest ancestors migrated from Africa, most probably in search of new resources of food as early as 80,000-70,000 years ago, there exist only tiny glimpses of what we were capable of achieving at this time.

 

For instance, around 80,000 years ago the peoples of the republic of Congo were making barbed bone hooks for fishing, while a community that inhabited a large cave called Blombos on the southern coast of South Africa would seem to have fashioned the earliest known examples of expressive art.

 

These take the form of incised pieces of red ochre rock, showing cross-hatching designs, as well as perforated snail shell beads, once strung on a cord and worn either as a necklace or bracelet.

 

All of these invaluable objects are thought to be around 75,000 years old.

 

 

Age of the Artist

In spite of the discovery of these clearly sophisticated personal items, whether practical or aesthetic, it was not until the start of the Upper Paleolithic age around 40,000 years ago that something quite dramatic began to occur.

 

At a time coincident to when homo sapiens first entered a Europe dominated until this time by the Neanderthal folk, there is clear evidence for the adoption of a complex life style, the earliest known to human kind.

It involved religious expression and practices, including detailed funerary rites, as well as magnificent new forms of art, such as the carving of animals, birds and humans in bone and stone and, crucially, the first appearance of cave art, such as the extraordinary painted galleries discovered as recently as 1994 at the Chauvet cave in southern France.

 

Dating to some 35,000 years before the present time, they contain images and sculptures of whole menageries of wild animals, including horses, rhinos, lions, mammoths and bison, alongside abstract representations of the human form extenuated, or brought to life, by the rock face itself.

Rapidly, hundreds of cave systems across Western Europe became full of sophisticated and highly accomplished art forms, a tradition which continued through until around 17,000 years ago, when suddenly there was a dramatic increase in sacred painting deep underground.

 

This trend ended finally around 11,000 years ago when the Upper Paleolithic age climaxed coincident to the cessation of the last Ice Age.

What Denis Montgomery wondered was whether that, in addition to other environmental, climatic and human factors, the increase in cosmic rays around 35,000 years ago, perhaps from the assumed supernova explosion which caused the creation of the Cygnus Veil, acted as a mutagen to effect sudden changes in the brain's neurological processes.

 

This in turn might have brought about the enlightened age of the cave artist in Western Europe. It could also explain why the Neanderthal peoples suddenly became extinct around this time, perhaps as a result of too much competition from their competitive new neighbors, the homo sapiens.

Montgomery's unique ideas were privately published, and, inevitably, largely ignored by the scholarly community (they can downloaded free from www.sondela.co.uk).

 

Adding to his problems was the realization by astronomers during the mid 1990s that the Cygnus Veil, the nebula at the centre of what Montgomery came to refer to as 'the Cygnus event', was found to be not 150 light years away from the Earth, as had previously been thought, but in fact around 1,800 light years away.

 

This meant that from here the supernova explosion would have been little more than a bright light source in the northern sky lasting for a period of several days, before gradually dying away.

 

Doubly damning were recalculations concerning the age of the supernova event, which now appears to have occurred as recently as 5,000-8000 years ago (even though some astronomical sources still reckon it took place much earlier, perhaps 10,000-15,000 years ago).

 

Thus there was no way that the Cygnus Veil can have been responsible for the high levels of cosmic rays reaching Earth's atmosphere when the first cave artists created underworld Sistern Chapels like Chauvet around 35,000 years ago.

 

Enter the Meinel Group

It would not be until 2005 that this same cosmological conundrum would be tackled again.

 

At the conference of TAG (the Theoretical Archaeological Group) in Sheffield, England, held in December that year, Dr Aden Meinel - a retired veteran of NASA's Jet Propulsion Laboratory, who in the 1980s was responsible for the launch of space telescopes such as Hubble - told a packed audience of bemused archaeologists and students that he and his colleagues at the Meinel Institute in Pasadena, California, had determined that the high levels of Beryllium 10 in the Greenland and Antarctica ice cores were responsible for sudden changes in evolution in both animal and human life around 40,000-35,000 years ago.

He also reported that he had been able to use the ice core evidence to determine the approximate coordinates for the source of the cosmic rays, and that these pinpointed a planetary nebula (a mass of glowing gas and cloud) known as the Cat's Eye in the northern constellation of Draco, the celestial dragon.

 

This the Meinel group saw as the remnants of what was once a galactic binary system consisting of a super giant and an active black hole spewing out jets of plasma, or ionized gas at velocities close to the speed of light.

 

These, he proposed, had crossed thousands of light years of space to reach the earth around 40,000 to 35,000 years ago, causing the changes in evolution witnessed at this time.

It was a dramatic claim, and one that needed scientific evaluation, which is where I entered this gripping story.

 

My own research into the emergence of primitive societies, with their own unique cosmologies and religion, ha revealed an inordinate interest in one particular constellation - Cygnus, the celestial swan. Indeed, it features as the oldest known artistic representation of a constellation anywhere in the world, for it is seen on the walls of the famous Lascaux cave in southern France, which is known to have first been occupied around 17,000 years ago.

Cygnus also appears as a bird in Church Hole cave in Derbyshire's Creswell Crags alongside cave art dated to 12,800 years ago, while at Göbekli Tepe in southeast Turkey, an 11,500-year-old stone temple - the oldest anywhere in the world - is orientated towards this same constellation.

 

It is the same story with ancient stone and earthen structures worldwide, from the bird effigy mounds of North America to,

• the Olmec centers of Mexico
• the Incan sacred city of Cuzco
• the Egyptian Pyramids of Giza
• the Hindu temples of India,

...to Avebury, the largest stone circle in Europe - all seem to reflect an age-old interest in Cygnus, which features also at the heart of religious symbolism worldwide, including that of Judaism, Christianity, Hinduism, Islam and more primitive forms of shamanism.

Putting aside more obvious astronomical reasons as to why our ancestors might have been interested in this particular constellation, shown universally as a celestial bird, I searched for answers as to why it might have been depicted deep underground by the cave artists of the Upper Paleolithic age.

 

In the knowledge that the work of South African anthropologist and rock art specialist David Lewis-Williams had determined that much prehistoric cave art was inspired by shamans in mind-altered states, I wondered whether the stars of Cygnus had come to be seen as the source of a primeval being, thought to have been responsible for cosmic life and death.

 

Moreover, did the peoples of the Upper Paleolithic come to associate this primeval cause with religious experiences deep underground, where their most sacred cave art was executed?

 

If so, then why did they come to associate a specific star constellation with the deepest part of caves?

 

The Soudan underground detector

in Minnesota

(Pic credit: Soudan)

I searched for answers, and was intrigued to find that in the early 1980s particle physicists using two newly-operational proton decay detectors - Soudan, located half a mile down an old mine in Minnesota, and NUSEX found off a road tunnel deep beneath France's Mont Blanc - detected the presence of highly unusual cosmic rays.

 

They bore the distinctive 'fingerprint' of a strange binary system containing a black hole or neutron star known as Cygnus X-3.

 

The existence of these high energy cosmic particles caused immediate controversy, since they matched the characteristics of no known particle.

Cygnus X-3 was identified as the first confirmed source of high energy cosmic rays able to penetrate deep underground, and then in 2000 NASA announced that it was the galaxy's first confirmed 'blazar', meaning that it was producing plasma jets that periodically sprayed the earth with volleys of cosmic rays, something it has been doing for anything up to 700,000 years.

The concerted reverence for Cygnus as the bird of creation and the source of cosmic life and death, led me to conclude that those responsible for the cave art at Lascaux, most probably shamans skilled in achieving altered states, had somehow become aware of incoming cosmic rays when deep underground.

 

Scientists have long been aware that when cosmic rays disintegrate they produce what is known as Cherenkov radiation, which appears as an objective burst of light as it passes through the water of the eye.

This phenomenon was first noted in 1968 when astronauts abroad Apollo 11 on its journey to the moon found that in the darkness they could see tiny flashes of light, either with their eyes open or closed. A NASA-funded investigation was quickly launched, which found that these light flashes were caused by cosmic rays passing through the Apollo spacecraft.

 

Did the Paleolithic shamans deep underground come to identify similar light bursts caused by cosmic rays from Cygnus X-3 as part of some religious experience, centered on the Cygnus constellation?

 

Moreover, did the high levels of cosmic rays that affected the world in Paleolithic times come from the direction of Cygnus, explaining why it can be found at the heart of religious symbolism worldwide?

 

Cygnus X-3 taken by the Chandra X-ray observatory

 (Pic credit NASA/Chandra X-ray Observatory ACIS/HETG).

Yet now Cygnus X-3 had a rival candidate in the Cat's Eye nebula, the chosen candidate for cosmic rays proposed by Aden Meinel and the Meinel Institute.

 

Unfortunately, astrophysicists are unanimous in their opinion that the Cat's Eye is a wimpy object unable to produce cosmic rays that might reach the Earth.

What is more, the Meinels, by their own admission, looked first in Cygnus for a possible point source of cosmic rays, and had found none - Cygnus X-3 being overlooked.

When in 1973 Carl Sagan wrote that cosmic rays might have been responsible for changes in human evolution he boldly asserted that their source was most probably a neutron star, which he saw as one of the most fascinating stellar bodies in the whole of the universe.

 

 

Artist's impression of

a black hole/neutron star binary system like Cygnus X-3

(Pic credit: NASA).

 

 

Today there can be little doubt that Cygnus X-3, as a neutron star/black hole as well as the galaxy's first confirmed blazar, is the best candidate by far for at least a proportion of the cosmic radiation responsible for the acceleration of human evolution at a time when we were just beginning to emerge as modern human beings.

Yet more disturbing is the fact that Cygnus X-3 is still out there, its cosmic gun barrel trained on the Earth, ever ready to release a volley of cosmic particles in our direction.

 

It has burst into action three times already this year, yet astronomers are waiting for what they see as the 'next big bang', showers of cosmic particles on a level never seen before, and when this happens, who knows, we might well be ready for the next stage in evolution.

Quote from smAsh on January 4, 2020, 8:54 am

More from Andrew Collins:

Cygnus X-3 and the Cosmic Ray Question

Andrew B Collins, science and history writer reports

Photograph of Cygnus X-3 taken by the Chandra X-ray Observatory in November 2000.
The horizontal line is thought to be an artefact of the image(Pic: (Credit: NASA/SRON/MPE).

 

Abstract

Cygnus X-3 is a high mass X-ray binary and microquasar, with a compact star, either a neutron star or a black hole, and a companion star, most probably a Wolf-Rayet (WN7 or 8) with huge mass loss and powerful stellar wind. First observed in 1966 (Giacconi, et al, 1967), Cygnus X-3 has been monitored across multiple frequencies, from radio, infrared, optical, X-ray to gamma-rays. It is one of the brightest galactic X-ray sources, and is the outright brightest during the production of bright radio flares, which can reach 20 Jy. Criticisms regarding Cygnus X-3 being a source of cosmic rays up to PeV (Meinels, personal communication, 11 July 2006) will necessitate a review of findings and theories since 1985. I will review recent evidence of Cygnus X-3's production of relativistic jets, as well as speculation that its core might be a source of strange quark matter, producing exotic primary particles, with the H being a possible candidate.

Sections: 1 Introduction; 2 Cygnus X-3 as a gamma-ray source; 3 Characteristics of Cygnet Primaries: 3.1 Production of muon flux deep underground; 4 Cygnet Identification; 5 Cygnets and strange quark matter; 6 Strange matter; 7 Cygnets, Neutrons and Relativistic Flow; 8 Some Criticisms of Cygnus X-3 as a Cosmic Accelerator; 9 Conclusions.

Key words: Cygnus X-3, gamma-rays, x-rays, excess muon fluxes, cygnets, neutrons, beta decay, neutral atoms, hadron-inducing, underground particle detection, strange quark matter, strange quarks, H particle, R-odd particle, particle acceleration, neutron star, strange star, linear acceleration, cosmic rays, relativistic jets, synchroton radiation, blazar, radio flares, jet velocity, SO particles.

1 Introduction
Cygnus X-3 (RA 307.6 dec 40.8) has been identified as a source of ultra high energy (UHE) gamma-rays of an extremely energetic nature. Indeed, their initial discovery in the 1970s was responsible for a complete reassessment of particle acceleration in compact stars. As early as 1973 the SAS-2 satellite reported gamma-radiation with a narrow phase interval of 4.8 h, noted separately in connection with x-ray and infrared observations of Cygnus X-3, estimated to be at 10 kiloparsecs (kpc, about 30,000 light years). This periodicity is most likely caused by the eclipsing of the compact star by its companion (Hillas, 1984), since jet precession is now calculated to be in the range of 5 days (Miller-Jones, et al 2004). Cygnus X-3 is also thought to be a sporadic 12.6 ms pulsar (Chadwick, 1985) with gamma-rays produced at or near the maximum (phase 0.6) in the 4.8 h X-ray cycle (Bowden et al, 1992).

2 Cygnus X-3 as a gamma-ray source
The gamma-ray emissions in association with Cygnus X-3 are known to range between 35 and 200 MeV, although the COS-B satellite between 1975 and 1982 reported no radiation between 70 and 5000 MeV with the point-source signature of 4.8 h (Cordova, 1986). Yet up to 1986 more than a dozen groups had reported the detection of gamma-rays from Cygnus X-3 with energies of at least 1011 eV. Gamma-rays in the higher range E> 1014 ev were subsequently reported and verified by ground-based collaborations in Germany, England, the United States, India and Italy (Cordova, 1986, and sources quoted).

The extremely energetic gamma-rays from Cygnus X-3 were early considered to be 'the products of interactions between even more energetic particles within the source, mainly protons', leading to speculation that Cygnus X-3 was 'the first astronomical object to be identified with reasonable certainty as a source of cosmic rays', i.e. any cosmic radiation above 108 ev (Cordova, 1986), or, indeed, a 'cosmic accelerator' (Dar, 1986). Moreover, gamma-rays from Cygnus X-3 indicated that 'only a very small number of sources of like nature would be required to produce most of the observed high-energy cosmic rays.'(Cordova, 1986).

Among the suspected method of production of gamma-rays were two popular models. Either they were protons accelerated by the electric field induced in the accretion disk held in the magnetic field of the compact star, or they were accelerated by shocks in the matter accreted on to a neutron star or black hole.

3 Characteristics of Cygnet Primaries
Between 1983 and 30 October 1985 various ground-based air shower arrays, including Kiel (Samorski and Stamm, 1983a, 1983b) and Fly's Eye (the latter from 1981 through till 1988) reported extensive air showers with the direction and periodicity of Cygnus X-3 (See Marshak et al, 1985; Cassiday et al, 1989). In Kiel's case, particles were detected in the 1016 eV range (initially assumed to be gamma-rays). This was later confirmed (Lloyd-Evans et al, 1983) with the pulse being narrow (duty cycle 2%) and occurring at a phase 0.25 after the X-ray maximum. Thus it was concluded that Cygnus X-3 accelerated particles to at least 1016 eV, and that if these were electrons, then protons might reach a higher level still (Hillas, 1984). Indeed, at Kiel the EAS reached energies of > 1018 eV (Cassiday et al, 1989; Sommer and Elbert, 1990).

At the same time two underground nucleon-decay detectors set up originally to observe proton decay, Soudon (Marshak et al, 1985) and NUSEX (Battistoni, 1985, Baym, 1985), reported excessive muon fluxes either with a time modulation of the 4.8-h period of Cygnus X-3, or coincident to its daily transit. The flux from single-muon events was greater than several orders than that expected from high energy photon flux, suggesting most probably either a primary of unique characteristics, dubbed the 'cygnet', or a new mechanism for very efficient muon production in high energy photon-initiated air cascades (Dar, 1986).

3.1 Production of Muon Flux Deep Underground
Excess muon quanta reported deep underground (at a depth equivalent of 2 to 5 kilometres of water) produced an angular spread highly suggestive of the primaries interacting in the rock overhang down to a depth of a few hundred metres (Kolb, 1985, Ruddick, 1986). This was confirmed by the differences in flux between the Soudon and NUSEX detectors, with the latter's flux being ten times less than the former, leading to the conclusion that this effect 'can only be explained by attenuation of the cygnet beam in the rock' (Ruddick, 1986). It also explained the zenith angles of the muons, which were similar to the background flux produced by EAS. Furthermore, the underground muon energy measurements predicted a characteristic variation of the quanta according to depth, with detectors on the surface only being able to detect them at near the horizontal, due to the large interaction length of the primaries. This meant 'such detectors would have to be very large to detect a signal'(Ruddick, 1986).

4 Cygnet Identification
Identification of the relativistic primaries (Ruddick, 1986) responsible for these signal events has proved extremely difficult, highly controversial and even questionable (Thomsen, 1986). The enhanced muon flux recorded, particularly by the Soudon I group, was far too high for them to be gamma-rays, which produced a mere 1/300 of the muons (µ-mesons) characteristic of the reported muon excess (Baym, 1986). Their 4.8-hour periodicity meant that they had to have travelled in a more or less straight line at relativistic speeds, otherwise a spread of lower velocities would have washed out the signal. Clearly, the path of the cygnets was not curved by the galactic magnetic field, otherwise this would have randomised or deflected their arrival directions (Dar, Lord and Wilkes, 1986).

The fact that the cygnets produced excessive muon (µ-mesons) quanta, implied that they acted to produce hadron-induced cascades. In other words, they were strongly-interacting particles, rather than electromagnetic particles, like gamma-rays or weak particles such as neutrinos (Dar, Lord and Wilkes, 1986). Further evidence against them being neutrinos was the fact that the cygnet-produced muon flux increased when Cygnus X-3 was overhead, and faded when it was not in view, the so-called 'horizon effect'. This is not a characteristic of neutrinos, which do not interact in this manner.

Thus the conclusion was that cygnet primaries, either measured in ground-based air shower arrays or in underground detectors, were long-lived neutral particles with energies anything up to at least PeV (Kolb, 1985; Maiani, 1985; Berezinsky, Ellis and Ioffe, 1986; Cordova, 1986; Ruddick, 1986; Cassiday et al, 1989; Czapek et al, 1990). However, the only obvious candidate was the neutron, which is unstable to beta decay and has a half-life of approximately 10-15 minutes. Thus the only way that they could have reached the earth was by travelling at relativistic speeds. Quashing this possibility was the fact that it would require neutrons with 100 times the energy of the monitored cygnet events. Neutral atoms could be eliminated since their electrons would have been stripped away by the 5g per cm2 interstellar hydrogen, causing their decay long before they ever reached the Earth. This is unless they had an incredibly high energy in the range of 1018 ev (Baym, 1986). As already noted, the Kiel collaboration did indeed register EAS with energies >1018 eV.

Initial findings strongly indicated that the cygnet primary bore the following characteristics: 1) no electric charge; 2) no magnetic charge; 3) a rest mass estimated to be between zero and 1/20 of a proton mass, and less than its energy by a factor of around 104 (Baym, 1986; Ruddick, 1986); 4) it was strongly interacting, in that it was hadron-inducing, and, lastly; 5) it possessed a half-life relative to its assumed passage at relativistic speed. Protons, neutrons, nuclei, atoms, and micrograins of ordinary matter could all be ruled out.

The cygnets were not charged cosmic particles since they are affected by the galactic magnetic field which randomizes their directional flow and ruins any chances of ascertaining their astronomical source, which can only be determined if they correlate with activity in other frequencies that might contain a known periodicity or direction.

Since neutral primaries arrive directly from source without being affected by the galactic magnetic field, they are crucial to determining the original trajectory of any cosmic ray. Gamma-rays are neutral, and so can also arrive directly from source, why we can trace the astronomical source of GRBs.

Thus in order to determine point sources of cosmic rays it is better to examine neutral particles, which retain their primary trajectory across the galaxy and when travelling at relativistic speeds will also retain their unique signature, which has been the case with Cygnus X-3 and Hercules X-1.

Indeed, as long ago as 1983 it was suggested that 'since the galactic magnetic field seems sufficient to randomize all charged particles during their long flights through space, pristine cosmic rays may not be charged particles at all.' (See Hecht and Torrey, 1983). A study of the five strongest recorded UHE cosmic ray events (E>1020 eV) by Farrar and Torrey (1998) led to the conclusion that their trajectory pointed back to extra-galactic QSOs (quasar stellar objects) with a margin of error of 0.005. In order that the primaries do not violate the Greisen-Zatsepin-Kuzmin (GZK) cutoff for distance travel of a photon or nuclei, they saw them as long-living, neutral hadrons of a possible exotic nature (Farrar, et al, 1998).

5 Cygnets and Strange Quark Matter
As far back as the mid 1980s it was proposed that cygnets were exotic hadrons resulting from strange quarks produced in strange quark matter in the core of Cygnus X-3. Maiani (1985), for instance, noted the observation of 'very energetic particles' arriving from Cygnus X-3, with estimated energies up to 104 TeV, as well as the observation underground of high energy muons correlated with a 4.8h modulation. He also accepted that poorer statistics might have been behind why other collaborations failed to register these increased muon fluxes underground, such as FREJUS and HPW. The primaries, he suggested, could be photons, which produce high energy showers in the atmosphere, and might well explain underground muon fluxes like those observed by Soudon and NUSEX. Yet results from these collaborations showed an increased muon flux too high for photons to be the simple answer (Dar, 1986). In contrast to Kolb and Ruddick, Maiani saw an anticorrelation in the reported muon flux versus the depth of the traversed rock, and the fact that NUSEX results were less than Soudon I. This was evidence, he felt, merely of the 'absorption effect' (Maiani, 1985).

6 Strange Matter
Kolb (1985) asserted that quark nuggets might lead to an enhancement in muon production over normal nuclear matter, yet even then only by a factor of two. He additionally considered the R-odd particle from supersymmetry and also the H-particle, after Jaffe. This last cited he saw as having a lifetime long enough to reach the earth, because of its double beta decay. Moreover, it bears four of the main characteristics of the proposed cygnet, although whether it can produce the enhanced muon flux depended upon its method of production. Despite this, the mass of the H, might not be low enough, something that only experimentation could determine. Even if the mass was close to that of the cygnet, the muon flux would be smaller than that reported. Moreover, the H cannot account for the angular spread of the underground muon flux. For instance, the NUSEX signal was seen coming from a 10 degree by 10 degree window in celestial coordinates, larger than the 0.5 degrees expected for an angular resolution.

Baym (1986) likewise proposed that the cygnet primary was the theorised H particle. Should its mass be less than that of the lambda (?) (1.1 16 GeV) plus that of the neutron (0.938 GeV), then it was possible that the lifetime of the H could be sufficiently long for it to be the cygnet primary, since it would not undergo the rapid decay into a single lambda or neutron.

Wilk and Wlodarczyk (1996) acknowledged 'anomalous cosmic ray bursts from Cygnus X-3' as the result of strange quark matter existing in its presumed neutron star (Wilk and Wlodarczyk, 1996). This supposition was explored further by Rybczynski, Wlodarczyk and Wilk (2004).

In similar to Kolb and Baym, Weber (2005) saw further support for the existence of strange matter in Cygnus X-3, which he speculates produces cosmic rays that to arrive as point-source signal events means that they have to be 'electrically neutral', like the cygnet primaries. Acknowledging their main characteristics, he confirms also that to survive the trip from Cygnus X-3 such particles are 'long-lived'. In his opinion, the 'only known particle which can quantitatively satisfy this constraint is the photon'. This is despite the fact that, as he states, they would only produce air showers with a 'small muon component'.

Weber goes on to predict that the 'natural candidate' for the cygnet is the H particle. Their potentially long lifetime means that they 'may be present as components of existing neutral particle beams' (Weber, 2005). Yet in order to give it long life, it would need to have 'mass below single weak decay stability'. Furthermore, in order to generate enough H particles, the source would have to be a strange star. Weber admits that the problem with Cygnus X-3 is that, 'it is accreting mass and thus has a crust, such that there is no exposed strange matter surface where small strangelets could be produced and subsequently accelerated electrodynamically to high energies into the atmosphere of the companion star where H particles were created via spallation reactions.'(Weber, 2005). Yet other evidence of strangelets in balloon-borne counter experiments, air-shower arrays and large emulsion chambers has convinced Weber that 'some primary cosmic rays may contain non-nucleus components which generate extended air showers that contain both a large number of muons as well as very high energetic photons', with Cygnus X-3 being a unique candidate.

7 Cygnets, Neutrons and Relativistic Flow
The idea that cygnet primaries are the result of exotic nuclei within Cygnus X-3's compact star being accelerated towards the earth is based on the view that they interact as hadrons to induce cascades uncommon to the normal production of muons in the atmosphere. However, should it be shown that the particles travel at relativistic speeds and thus contain considerably higher energies than previously reported, then they might be explainable in more conventional terms. Soudon reported that the muon excess from Cygnus X-3 was coincident to major radio flares (from 0.1 mJy up to 20Jy), which have themselves occurred coincident with X-ray and infrared observations (Baym, 1986). Moreover, gamma-rays with the 4.8 h periodicity of Cygnus X-3 monitored by ground-based air arrays have also coincided with considerable excess muon flux underground as mentioned earlier.

Sommers and Elbert (1989) examined the evidence for EeV neutrons and/or photons from Cygnus X-3, based on the Fly's Eye data, and stated that 'because of synchroton radiation losses, EeV particle acceleration cannot occur gradually while a particle orbits in a strong magnetic field.' As a consequence, Sommers and Elbert suspected that 'if particles are accelerated in a neutron star's magnetosphere, some type of linear accelerator must be responsible (my italics)'.

Accepting that the cosmic rays from Cygnus X-3 are neutral, since charged particles would be dispersed by the galactic magnetic field at EeV energies, the question remains of how neutral particles might be produced by accelerated charged particles. According to Sommers and Elbert, the range of possible models for the production of EeV neutral particles 'is greater than the range considered for the production of 1015 eV neutral particles from Cyg X-3. This is partly because the EeV neutral particles can be neutrons as well as gamma-rays' (Sommers and Elbert, 1989).

Crucially, Sommers and Elbert go on to state that 'although free neutrons decay with a mean proper lifetime of 898 seconds', time dilation allows some neutrons at these energies to travel the distance from Cyg X-3. On this basis, the energy threshold (0.5 EeV) for the data used in the Fly's Eye analysis suggests that the reported increased muon flux could be neutrons, even though the collaboration was at the time unable to distinguish between a neutron-initiated shower and a gamma-ray shower (Sommers and Elbert, 1989). In final conclusion, they stated that 'Cyg X-3 is a strong source of EeV cosmic rays'.
The significance of Sommers and Elbert's proposal is that with a relativistic linear acceleration through jet production, the view that cygnets are exotic strange quark particles becomes unnecessary. The neutral particles might indeed be neutrons, reliant on a new model based upon synchrotron radiation loss through relativistic flow.

Cygnus X-3 during its major radio flare in September 2001 (after Miller Jones, et al, 2004).

A one-side relativistic jet was observed in association with radio flare activity in Cygnus X-3 by Mioduszewski, et al (2001) using the VLBA in February 1997. It was estimated to have an opening angle of 12 degrees, and a small inclination angle of > 12 degrees towards the Earth, leading to the conclusion that it constituted the galaxy's first blazar. A precessional phase of 30 days with an anticlockwise movement was also noted in association with jet production. These parameters were comparable with those obtained during the observation in September 2001 of a separate major radio flare by Miller-Jones, et al. (2004) using the VLBA over a period of six days following a peak outburst. The southern jet was estimated to be moving within 10.5 degrees to the line of sight, with a precessional phase in a clockwise motion of 5.3 days. The northern jet was weakly observed. Clearly, the implication was that both the precession cycle and the direction of the jets had shifted between 1997 and 2001. Through extrapolation of the jet motion back to source Miller-Jones estimated that the jets were ejected about 2.5 days after the radio brightness of Cygnus X-3 began to increase. Overall the parameters of the southern jet have been found to be consistent with what Mioduszewski et al. had previously observed.

Bipolar jets were also observed in October-November 2000 using the VLA and examined by the NRAO (Marti, et al, 2002), although whether or not these were a separate mechanism to the observed north-south orientated jets observed in 2001 has still to be decided.

The speed of Cygnus X-3's suspected one-sided jet was originally estimated at 0.35c (Cordova, 1986). More recent assessments of the relativistic jets following the 1997 VLBA observations by Mioduszewski, et al (2001), provided a revised speed up to 0.81c, while Miller Jones, et al, following the 2001 observations concluded that the rate was 0.63c. Yet they accepted that faster speeds could precede the observable appearance of the series of bead-like knots marking the whereabouts of the jets; see also Hannikainen, et al (2003, poster) for further discussion on this subject.

Such speeds might be enough to enable short-lived neutrons to reach the Earth, indicating a realistic process for the arrival of low-mass neutral particles, and the possible production of increased air showers and underground muon quanta from Cygnus X-3.

It has been pointed out that an observed velocity of Cygnus X-3s jet at a maximum of 0.81c would be too slow to compress time so that neutrons might have time to decay into protons (Meinel, private communication, 11 July 2006).

The 0.81c for the speed of Cygnus X-3s southern jet for the February 1997 observations (Mioduszewski et al, 2001) is based on estimates of jet motion in radio flaring, and does not necessarily relate to the initial velocity of ejecta on all frequencies. Moreover, it is clear that the speed of the jets change, as is seen in the 2001 observations, where a velocity of just 0.63c was deduced (Miller-Jones, 2004). Earlier estimates of jet speed were even lower. Moreover, the bipolar jets monitored by the NRAO using the VLA in 2000 (Marti, et al, 2001) determined that they had an infrared speed of just 0.48c, which is much slower than the higher speed radio flares. Thus there is no reason why UHE and HE cosmic rays and gamma-rays from Cygnus X-3, or indeed any point source, might not exceed the velocity speed of radio flares.

Signal events from Cygnus X-3 which feature the arrival of GeV gamma-ray emissions and hadron-like neutral particles have coincided with intense bursts of energies across multiple frequencies during the production of jets. For instance, this occurred in October 1985 when an increased muon flux at PeV levels coincided with intense bursts of radio emissions (Berenzinsky, 1988). In addition to this, there was a correlation between the excess muon flux recorded by the Soudon II deep underground experiment between 1991 and 2000 and Cygnus X-3's production of large or intermediate radio flares (Thomson, 1991; Marshak, 2000; Allison, 1999). Thus there is every reason to conclude that the production of gamma-rays and long-lived neutral particles in Cygnus X-3 might well be the result of narrow, magnetically driven relativistic jets within a small angle of the Earth.

NRAO image showing the production of bipolar jets by Cygnus X-3 in October-November 2000 (pic: NRAO/AUI) .

8 Some Criticisms of Cygnus X-3 as a Cosmic Accelerator
It has been suggested that the cone angle of Cygnus X-3 poses a problem regarding any working model for it being a cosmic accelerator in its role as a blazar. In order to send cosmic rays in the Earth's direction with the solar region being significantly inside the limit to the probability distribution has been estimated to be about 0.5 radian (Meinel, private communication, 11 July 2006).

Work has yet to be undertaken on this level, and the recorded data from both particle physics and astrophysics needs careful evaluation in this respect. Much of the findings cited from the 1980s of long lasting neutral particles from Cygnus X-3 has largely been ignored. This is a shame, for it clearly suggests that Cygnus X-3 possesses an extraordinary acceleration mechanism for the production of cosmic rays. As we have seen , their appearance correlates well with radio flaring and hard X-ray outbursts. Nothing has so far been published on any possible correlations between cygnets and major radio flaring during the years 1997, 2000, 2001 and 2006.

The question with regards Cygnus X-3 is not whether it can accelerate out cosmic rays and UHE gamma-rays, but how exactly they might be produced. In my opinion, it is the production of relativistic jets, or shock waves in association with their production, that remain the best mechanism for their production, something predicted by Sommers and Elbert as far back as 1989. What is important about this and other similar conclusions during the 1980s is that there was no hard data available then suggesting that Cygnus X-3 might be a blazar. This came only through the February 1997 observations (see Mioduszewski, et al, 2001), and confirmed during the 2001 observations (Miller-Jones, et al, 2004). In other words, working models for the acceleration process of cosmic rays from Cygnus X-3 were shown to be correct, making the evidence of long-lasting neutral particles originating from here an extremely likely possibility. What is more, they have continued to be reported. Soudon II, announced in 1999 that in its first ten years of operation the collaboration had regularly tracked excess muon events in the Tev range or above from the direction of Cygnus X-3, and again in 2000 (Marshak, 2000). There is every likelihood that cygnets are accelerated from source during jet production, and that at such energies only stable, neutral particles can travel the 10 kpc distance from Cygnus X-3 to the earth 'along trajectories which point back to the source' (Allison, 1999). Moreover, the Soudon II collaboration conclude that since known stable neutral particles - photons and neutrinos - have only small probabilities of producing detectable muons 'Tev muons associated with Cygnus X-3 requires either exotic interactions of known primaries, exotic primaries or very large fluxes of neutrinos or photons' (Allison, 1999).

9 Conclusions
Despite the Soudon and NUSEX observations of an increased muon flux underground during the 1980s remaining controversial, there exists good evidence for the existence of long-lived, low-mass, strongly-interacting neutral primaries from Cygnus X-3.

Exotic primaries have been proposed to explain the reported EAS and increased muon flux underground, with strange quark matter from a strange matter compact star, most likely a neutron star, being currently the most popular model. However, with the introduction of a relativistic flow for the acceleration of the neutral primaries, it is possible that cygnet primaries are simply neutrons. Regardless of this, the idea of them being produced within neutron stars by strange matter remains an attractive theory, and exploration into this area of astrophysics is to be encouraged. It is ironic that the absence of a well-defined mechanism of production of these neutral primaries, along with their erratic data, has diminished the impact of these extremely important findings, which arguably hold the key to determining the first confirmed point source of galactic cosmic rays.

References
Allison, W W M, et al, 'Cygnus X-3 revisited: 10 years of muon and radio observations', Proceedings of the 26th International Cosmic Ray Conference (ICRC 99), Salt Lake City, Utah, 17-25 Aug 1999.
Battistoni, G, et al, 'Observation Of A Time Modulated Muon Flux In The Direction Of Cygnus X-3', 1985, Phys. Lett. 155B 465.
Baym, Gordon, 'Does Cygnus X-3 Contain a Strange Neutron Star?, Spring 1986, Los Alamos Science, 50-2.
Berezinsky, V S, Ellis, J, and Ioffe B L , 1986, Phys. Lett. 172B 423.
Berezinsky, V S, 'Time delay of the PeV gamma ray burst after the October 1985 radio flare of Cygnus X-3', 11 August 1988, Nature, 334, 506-7.
Bowden, C C G, et al, 'Recent observations of VHE gamma rays from Cygnus X-3', 1992, J. Phys. G: Nucl. Part. Phys. 18, 413-421.
Cassiday, G L, et al, 'Evidence for 1018-eV Neutral Particles from the Direction of Cygnus X-3', 23 January 1989, Phys. Rev. Lett. 62:4, 383-6.
Chadwick, P M, et al. 'A 12.6-ms pulsar in Cygnus X-3', Nature, 1985, 318, 642.
Cordova, France Anne-Dominic. 'Cygnus X-3 And The Case For Simultaneous Multifrequency Observations', Spring 1986, Los Alamos Science, 39-49, 53-4.
Dar, Arnon, Jere J Lord and R Jeffrey Wilkes, 'Nature of the high-energy particles from Cygnus X-3', Phys. Rev. D, 1 January 1986, D33:1, 303-6.
Farrar, Glennys R, and Biermann, Peter L, 'Correlation between Compact Radio Quasars and Ultrahigh Energy Cosmic Rays', 1998, Phys. Rev. Lett. 81, 3579-3582.
Giacconi, R, Gorenstein, P., Gursky, H., & Waters, J. R. 1967, ApJ 148, L119.
Hannikainen, D C, et al, 'Cygnus X-3: Multiwavelength monitoring during 1996-2002', 2003, poster.
Hecht, Jeff, and Torrey, Lee; 'Scientists Find Sources of Cosmic Rays,' New Scientist, 1983, 99:764.
Hillas, A M, 'The Origin of Ultra-High-Energy Cosmic Rays', 1984, Ann. Rev. Astrophys, 22, 425-44.
Kolb, E W, 'Searching For Cygnets', Sep 1985, Fermilab-conf-85-134-A, 1-12.
Lloyd-Evans, J, Coy, R N, Lambert, A, Lapikens, J, Patel, M, et al, 1983, Nature, 305, 784-7.
Maiani, Luciano, 'Electroweak Interactions', 1985, Conf. on High Energy Phys.
Marshak, M L, et al, ''Evidence for Muon Production by Particles from Cygnus X-3', 1985, Phys. Rev. Lett. 54, 2079-82.
Marshak, M L, 'Underground Muons Observed during the April 2000 Flare of Cygnus X-3', ICHEP 2000, ichep2000.hep.sci.osaka-u.ac.jp/abs_PA-11.html.
Marti, J, Paredes, J M, and Peracaula, M, 'Development of a two-sided Relativistic Jet in Cygnus X-3', 2001, A&A, 375, 476-84.
Miller-Jones, James CA, Blundell, Katherine M, et al, 'Time-sequenced MultiRadio Frequency Observations of Cygnus X-3 in Flare'AJ 600 (1 January 2004): 368-389
Mioduszewski, et al, 'Radio observations of Cygnus X-3 after the Large Radio Flare in September 2001', American Physical Society, April Meeting, Jointly Sponsored with the High Energy Astrophysics Division (HEAD) of the American Astronomical Society April 20 - 23, 2002 Albuquerque Convention Center Albuquerque, New Mexico Meeting ID: APR02, abstract #N17.089. www.nrao.edu/~amiodusz/Cygx3head02.pdf.
Mioduszewski, AJ, Rupen, MP, et al, 'A One-sided, Highly Relativistic Jet from Cygnus X-3', ApJ, 553 (2001), 766-775.
Ruddick, K, 'Model for the Underground Muons Associated with Cygnus X-3', 4 August 1986, Phys. Rev. Lett., 57: 5, 531-4.
Rybczynski, M, Wlodarczyk, Z, and Wilk, G, 'Strangelets in cosmic rays', 2002, Acta Phys. Polon. B33, 277-296.
Samorski, M, and Stamm W, 1983a, Astrophys. Jour. 268 L17.
Samorski, A, and Stamm W, 1983b, Proc. 18th Int. Conf. on Cosmic Rays, Bangalore EA1 1-39, ed N Durgaprased (Bombay: Tata Institute of Fundamental Research).
Sommers, Paul, and Elbert, Jerome W, 'Implications of the Fly's Eye Evidence for Neutral Particles from Cygnus X-3 with Energies Near 1018 eV', 1989, Astro. Lett. and Communications, 1990, 27, 397-402.
Thomsen, D E, 'A disbelief in Cygnets - Cygnus X3', SN 130:6 (9 August 1986), p.89.
Thomson, M A, et al, 'The Observation of Underground Muons from the Direction of Cygnus X-3 during the January 1991 Radio Flare', PRL B 269 (1991), pp. 220-6.
Weber, F, 'Strange Quark Matter and Compact Stars', 2005, Prog. Part. Nucl. Phys., 54, 193-288 (astro-ph/0407155).
Wilk, G, and Z Wlodarczyk, 'How Can Strange Quark Matter Occur Deeply in the Atmosphere?', 1996, J. Phys. G., G 22, L105-L111.

 

Quote from smAsh on January 5, 2020, 7:18 am

In searching for a different document, I found this one.  Standard issue science, here.  Nearly every time and answer is offered by nature, countless questions arise...

Photoionization Emission Models for the Cyg X-3 X-Ray Spectrum-

Abstract:
We present model fits to the X-ray line spectrum of the well-known high-mass X-ray binary Cyg X-3. The primary
observational data set is a spectrum taken with the Chandra X-ray Observatory High Energy Transmission Grating
in 2006, though we compare it to all the other observations of this source taken so far by this instrument. We show
that the density must be …1012 cm−3 in the region responsible for most of the emission. We discuss the influence of
the dust scattering halo on the broadband spectrum, and we argue that dust scattering and extinction is not the most
likely origin for the narrow feature seen near the Si K edge. We identify the features of a wind in the profiles of the
strong resonance lines and show that the wind is more apparent in the lines from the lighter elements. We argue
that this wind is most likely associated with the companion star. We show that the intensities of most lines can be
fitted, crudely, by a single-component photoionized model. However, the iron K lines do not fit with this model.
We show that the iron K line variability as a function of orbital phase is different from the lower-energy lines,
which indicates that the lines arise in physically distinct regions. We discuss the interpretation of these results in
the context of what is known about the system and similar systems.
Key words: stars: black holes – stars: winds, outflows – X-rays: binaries

Here's a link to the paper from 2019!

https://iopscience.iop.org/article/10.3847/1538-4357/ab09f8/pdf

 

 

Quote from smAsh on January 5, 2020, 7:22 am

And HERES ANOTHER from 2000:

https://iopscience.iop.org/article/10.1086/312608/fulltext/

 

High-Resolution Spectroscopy of the X-Rayphotoionized Wind in Cygnus X-3 with the Chandra High-Energy Transmission Grating Spectrometer

Frits Paerels ,^1,2 Jean Cottam ,¹ Masao Sako ,¹ Duane A. Liedahl ,³ A. C. Brinkman ,² R. L. J. van der Meer ,² J. S. Kaastra ,² and P. Predehl ⁴

Received 2000 January 20; accepted 2000 February 11; published 2000 March 24

ABSTRACT

We present a preliminary analysis of the 110 keV spectrum of the massive X-ray binary Cygnus X-3, obtained with the high-energy transmission grating spectrometer on the Chandra X-Ray Observatory. The source reveals a richly detailed discrete emission spectrum, with clear signatures of photoionization-driven excitation. Among the spectroscopic novelties in the data are the first astrophysical detections of a number of He-like "triplets" (Si, S, Ar) with emission-line ratios characteristic of photoionization equilibrium, fully resolved narrow radiative recombination continua of Mg, Si, and S, the presence of the H-like Fe Balmer series, and a clear detection of an 800 km s^-1 large-scale velocity field as well as an 1500 km s^-1 FWHM Doppler broadening in the source. We briefly touch on the implications of these findings for the structure of the Wolf-Rayet wind.

Subject headings: atomic processes; stars: individual (Cygnus X-3); techniques: spectroscopic; X-rays: stars

     ¹ Columbia Astrophysics Laboratory, Columbia University, 538 West 120th Street, New York, NY 10027.
     ² Space Research Organization of the Netherlands, Laboratory for Space Research, Sorbonnelaan 2, Utrecht, CA NL-3584, Netherlands.
     ³ Department of Physics, Lawrence Livermore National Laboratory, P.O. Box 808, L-41, Livermore, CA 94550.
     ⁴ Max-Planck-Institut für extraterrestrische Physik, Giessenbachstrasse, Postfach 1603, Garching, D-85740, Germany.

1.  INTRODUCTION

In a previous Letter (Liedahl & Paerels 1996, hereafter LP96), we presented an interpretation of the discrete spectrum of Cygnus X-3 as observed with the solid-state imaging spectrometers on ASCA (Kitamoto et al. 1994; Kawashima & Kitamoto 1996). We found clear spectroscopic evidence that the discrete emission is excited by recombination in a tenuous X-rayphotoionized medium, presumably the stellar wind from the Wolf-Rayet companion star (van Kerkwijk et al. 1992). Specifically, the ASCA spectrum revealed a narrow radiative recombination continuum (RRC) from H-like S, unblended with any other transitions. On closer inspection, RRC features due to H-like Mg and Si were also found to be present in the data, although severely blended with emission lines. These narrow continua are an unambiguous indicator of excitation by recombination in X-rayphotoionized gas, and their relative narrowness is a direct consequence of the fact that a highly ionized photoionized plasma is generally much cooler than a collisionally ionized plasma of comparable mean ionization (see LP96, Liedahl 1999, and references therein).

With the high spectral resolution of the Chandra high-energy transmission grating spectrometer, we now have the capability to fully resolve the discrete spectrum. Apart from offering a unique way to determine the structure of the wind of a massive star, study of the spectrum may yield other significant benefits. Cyg X-3 shows a bright, purely photoionization-driven spectrum and, as such, may provide a template for the study of the spectra of more complex accretion-driven sources, such as active galactic nuclei. The analysis will also allow us to verify explicitly the predictions for the structure of X-rayphotoionized nebulae derived from widely applied X-ray photoionization codes.

2.  DATA REDUCTION

A description of the high-energy transmission grating spectrometer (HETGS) may be found in Markert et al. (1994). Cyg X-3 was observed on 1999 October 20 for a total of 14.6 ks exposure time, starting at 01:11:38 UT. The observation covered approximate binary phases -0.31 to +0.53, which means that about half of the exposure in our observation occurs in the broad minimum in the light curve at orbital phase zero. Aspect-corrected data from the standard CXC pipeline (processing date: 1999 October 30) was postprocessed using dedicated procedures written at Columbia University. We used (ASCA) grade 0, 2, 3, and 4 events; a spatial filter 30 ACIS pixels wide was applied to both the high-energy grating (HEG) and medium-energy grating (MEG) spectra, and the resulting events were plotted in a dispersionCCD pulse-height diagram, in which the spectral orders are neatly separated.

A second filter was applied in this dispersionpulse-height diagram. The filter consisted of a narrow mask centered on each of the spectral orders separately. The mask size and shape were optimized interactively. The residual background in the extracted spectra is of order 0.5 counts per spectral bin of 0.005 Å or less. The current state of the calibration does not provide us with the effective area associated with our joint spatial/pulse height filters to better than 25% accuracy; hence, we have chosen not to flux-calibrate the spectrum at this time. An additional correction to the flux in the chosen aperture due to the (energy-dependent) scattering of photons by interstellar dust has not yet been determined either.

In the resulting order-separated count spectra, we located the zero order, and we determined its centroid position to find the zero of the wavelength scale. We then converted pixel number to wavelength based on the geometry of the HETGS. In this procedure, we used ACIS/S chip positions that were determined after launch from an analysis of the dispersion angles in the HETGS spectrum of Capella (Canizares et al. 2000). This preliminary wavelength scale appears to be accurate to approximately 2 mÅ. The spectral resolution was determined from a study of narrow, unblended emission lines in the spectrum of Capella. It is approximately constant across the entire HETGS band and amounts to approximately 0.012 Å (0.023 Å) FWHM for the HEG (MEG) (D. Dewey 2000, private communication). The resolution in the Cyg X-3 spectrum can be checked self-consistently by analyzing the width of the zero-order image. Unfortunately, the zero-order image is affected by pileup. However, enough events arrive during the 41 ms CCD frame transfer, forming a streak in the image, that we can construct an unbiased one-dimensional zero-order distribution from them. The width of this distribution is consistent with the widths of narrow lines in the spectrum of Capella, which indicates that the resolution in the Cyg X-3 spectrum is not affected by systematic effects (e.g., incorrect aspect solution, defocusing).

3.  X-RAY PHOTOIONIZATION IN CYG X-3

Figure 1 shows the HEG and MEG first-order spectra; the higher order spectra are unfortunately very weak, and we will not discuss them here. We show the spectra as a function of wavelength because this is the most natural unit for a diffractive spectrometer: the instruments have approximately constant wavelength resolution. The spectra have been smoothed with a 3 pixel boxcar average to bring out coherent features. We have indicated the positions of expected strong H- and He-like discrete features. A cursory examination of the spectrum strikingly confirms the photoionization-driven origin of the discrete emission.

Fig. 1   The 110 Å spectrum of Cyg X-3 as observed with the HEG (top) and the MEG (bottom) binned in 0.005 Å bins. The positive and negative first orders have been added, and the spectra have been smoothed with a 3 pixel boxcar filter. The labels indicate the positions of various discrete spectral features. "He" is the inelegant label for the resonance, intercombination, and forbidden lines in the He-like ions, plotted at the average wavelength for the complex. High-ionization features of interest that were not detected have been labeled in parentheses. The horizontal bars indicate the nominal positions of the gaps between the ACIS chips; the dithering of the spacecraft will broaden the gaps and soften their edges.

We detect the spectra of the H-like species of all abundant elements from Mg through Fe. In Si and S, we detect well-resolved narrow radiative recombination continua. This is illustrated in Figure 2, which shows the 3.07.0 Å band on an enlarged scale. The Si XIV and S XVI continua are readily apparent. The width of these features is a direct measure of the electron temperature in the recombining plasma, and a simple eyeball fit to the shapes indicates kT_e  50 eV, which is roughly in agreement with the result of model calculations for optically thin X-rayphotoionized nebulae (Kallman & McCray 1982). A more detailed, fully quantitative analysis of the spectrum will be required to see whether we can also detect the expected temperature gradient in the source (more highly ionized zones are also expected to be hotter). In the Si XIV and S XVI spectra, we estimate the ratio between the total photon flux in the RRC to that in Ly to be about 0.8 and 0.7, respectively; here we assume kT_e = 50 eV, and we have made an approximate correction for the differences in effective area at the various features. These measured ratios are in reasonable agreement with the expected ratio of 0.73(kT_e/20 eV)^+0.17 (LP96), which indicates that the H-like spectra are consistent with pure recombination in optically thin gas.

Fig. 2   The 3.07.0 Å region of the spectrum enlarged; we show the raw count rates, binned by two 0.005 Å bins. The most important transitions have been labeled; dotted lines mark the expected positions of Si and S recombination edges. These markers have been redshifted by 800 km s-1. The horizontal bar near 4.5 Å in the HEG spectrum marks the nominal position of the gap between chips S2 and S3 in ACIS. The solid line in the MEG spectrum is a crude empirical fit to the continuum, with Si XIII, Si XIV, and S XVI narrow radiative recombination continua added. The electron temperature was set to 50 eV, and the continua were convolved with a 1500 km s-1 FWHM velocity field, to match the broadening observed in the emission lines.

The positions of the lowest members of the Fe XXVI Balmer series are indicated in Figure 1 (the fine-structure splitting of these transitions is appreciable in H-like Fe, as is evident from the plot). The relative brightness of the Balmer spectrum is yet another indication of recombination excitation. There is evidence for line emission at the position of H, and possibly at H and H; the spectrum is unfortunately too heavily absorbed to permit a detection of H (9.52, 9.74). Unfortunately, the long-wavelength member of the H "doublet" (  7.17 Å) almost precisely coincides with the expected position of Al XIII Ly, which precludes a simple and neat direct detection of Al (the first detection of an odd-Z triple- element in nonsolar X-ray astronomy). Any limit on the Al/Si abundance ratio thus becomes dependent on an understanding of the intensity of the Fe XXVI spectrum.

As for the He-like species, we detect the n = 21 complexes, consisting of the forbidden (f), intercombination (i), and resonance (r) transitions, in Si XIII, S XV, Ar XVII, Ca XIX, and Fe XXV (as well as the corresponding RRC in Si, S, and possibly Ar). The line complexes appear resolved into blended resonance plus intercombination lines, and the forbidden line (see Figs. 1 and 2), up to Ar XVII.

In an optically thin, low-density, purely photoionization-driven plasma, one expects the intensity ratio f/(r + i)  1 for the mid-Z elements, very different from the pattern in the more familiar collisional equilibrium case, where the resonance transition is relatively much brighter (e.g., Gabriel & Jordan 1969; Pradhan 1982; Liedahl 1999). We use the ratio f/(r + i) rather than the conventional G  (i + f)/r and R  f/i because the intercombination and resonance lines are unfortunately blended by significant Doppler broadening in the source (see § 4). Theoretically, in a photoionized plasma, f/(r + i) is approximately equal to 1.3, 1.0, and 0.83 for Si XIII, S XV, and Ar XVII, respectively, and depends only weakly on electron temperature (LP96; D. A. Liedahl 2000, in preparation). The measured ratios f/(r + i), derived by fitting three Gaussians with common wavelength offset and broadening at the expected positions of f, i, and r, are approximately 1.1, 0.8, and 1.1 with the HEG for Si, S, and Ar, respectively; the corresponding ratios for the MEG are 1.3, 1.0, and 0.8. Since most of the lines contain at least 100 photons, the statistical error on the ratios is generally less than 15%. These measurements include a model for the Si XIII RRC in the S XV triplet (assuming kT_e = 50 eV) and Mg XII Ly emission in the Si XIII triplet.

The He-like line ratios are probably affected by systematic features in the efficiency of the spectrometer. The S XV triplet is superposed on the Si XIII RRC, the Si XIII triplet straddles the Si K edge in the CCD efficiency, and the Ar XVII triplet straddles the Au M_IV and Ir M_I edges. Corrections for these effects will have to be carefully evaluated. Nevertheless, the raw ratios f/(r + i) for the Si and Ar triplets are already of the right magnitude for pure recombination. Our provisional conclusion is that the He-like spectra are, very roughly, consistent with pure recombination in optically thin gas.

Just as in a collisional plasma, the relative strengths of the forbidden and intercombination lines are sensitive to density (Liedahl 1999; Porquet & Dubau 2000) because of the collisional transfer between the upper levels of f and i at high density. As mentioned above, there are some systematic uncertainties in the measured line ratios, and we defer a discussion of possible constraints on the density in the wind to a future paper.

The detection of fluorescent Fe emission is a surprise because virtually no fluorescence was seen at the time of the ASCA observation (Kitamoto et al. 1994). The apparent centroid wavelength of the fluorescent line is 1.939 Å (photon energy 6394 eV), with a formal error of less than 10^-3 Å (3 eV). The width of the line is 0.022 Å FWHM, with a formal uncertainty of less than 5%. This is wider than would be expected from the velocity broadening to be discussed in the next section and may be an indication that a range of ionization stages contributes to the fluorescent emission. If we assume the same velocity broadening for the Fe K feature as for the high-ionization lines (which may not necessarily be correct if the low- and high-ionization lines originate in different parts of the stellar wind), we find that Fe K has an intrinsic width (expressed as the FWHM of a Gaussian distribution) of 0.018 Å (corresponding to E  60 eV). The fine-structure split between K₁ and K₂ contributes slightly to this width (  0.004 Å), but the measured width covers the full range of K wavelengths for charge states between fully neutral and Ne-like (Decaux et al. 1995).

4.  BULK VELOCITY FIELDS

We find that all emission features are significantly broadened and redshifted. The lines and radiative recombination continua are resolved by both the HEG and the MEG. The line widths for H-like Mg, Si, S, Ar, Ca, and Fe Ly were measured by fitting a simple Gaussian profile. Other than the negligibly small fine-structure split (  0.005 Å), these lines are clean and unblended. The resulting widths do not seem to exhibit a strong dependence on phase. Assuming that the spectrometer profile is well represented by a Gaussian of width 0.012 Å (0.023 Å) FWHM for the HEG (MEG), we find that the broadening of the lines is roughly consistent with a Gaussian velocity distribution, of width v  1500 km s^-1 FWHM. The scatter is too large to permit a meaningful test for any dependence of the velocity broadening on ionization parameter. Note that no such broadening was seen in the spectrum of Capella.

We also measured the radial velocities for the Ly lines, assuming the dispersion relation obtained from an analysis of the spectrum of Capella. Wavelengths were calculated from the level energies given by Johnson & Soff (1985); these should be accurate to a few parts in 10⁶. There is a clear systematic redshift to all the emission lines and RRCs, in both the positive and negative spectral orders and in both grating spectra. This is shown in Figure 3, where we have segregated dim and bright state data but have averaged positive and negative spectral orders and HEG and MEG spectral data. Also shown are the best-fitting uniform velocity offsets. These fits were forced to yield zero wavelength shift at zero wavelength. The average redshift for the dim state is 800 km s^-1, and for the bright state, it is 750 km s^-1. We thus find a net redshift much smaller than the observed velocity spread and essentially no dependence of the centroid velocity on the binary phase. We should point out that our preliminary analysis, based on fitting simple Gaussians, is admittedly crude and may have biased the true nature of the velocity field somewhat. We also note, with caution, that Doppler shifts due to a single, uniform velocity do not appear to be a very good description of the data: the longest wavelength lines appear to be offset at a significantly larger-than-average radial velocity. A detailed analysis, taking into account the actual line shape, will be required to confirm or refute the possibility that these offsets represent the expected systematic correlation of average wind velocity and ionization parameter.

Fig. 3   Measured wavelength shift for selected Ly features. The filled squares refer to the "dim" state data, the open squares to the "bright" state data. The velocities as measured with the HEG and the MEG have been averaged; velocities in positive and negative spectral orders were averaged. Error bars indicate the size of the rms variation between these various measurements. In cases in which only one or two velocities were measurable because of a low signal-to-noise ratio, we instead indicate the estimated statistical error on these measurements. The solid lines are the weighted least-squares Doppler velocities for both the dim and the bright states.

5.  DISCUSSION

The HETGS spectrum of Cyg X-3 has revealed a rich discrete spectrum, the properties of which are consistent with pure recombination excitation in cool, optically thin, low-density X-rayphotoionized gas in equilibrium. We fully resolve the narrow RRCs for the first time and estimate an average electron temperature in the photoionized region of kT_e  50 eV, consistent with global photoionization calculations.

We detect a net redshift in the emission lines of v  750800 km s^-1, essentially independent of binary phase, and a distribution in velocity with an FWHM of 1500 km s^-1. If the wind were photoionized throughout, we would expect to see roughly equal amounts of blue- and redshifted material, so evidently we are viewing an ionized region that is not symmetric with respect to the source of the wind, as expected if only the part of the wind in the vicinity of the X-ray continuum source is ionized. However, in the simplest wind models, one would then expect to see a strong dependence of the centroid velocity on binary phase, alternating between red- and blueshifts, and this is decidedly not the case in our data. The implications of this finding for the flow pattern and distribution of material in the wind will be explored in a future paper.

Finally, the Fe K fluorescent feature, which probes a more neutral phase of the wind, has never been seen before in Cyg X-3. Unfortunately, the exact range of ionization cannot be separated uniquely from systematic Doppler shifts through a measurement of the wavelengths of the K spectra because the feature, while clearly broadened, is not separated into its component ionization stages. Still, the width of the feature (the net effect of the velocity field and the existence of a range of charge states) and its intensity will impose strong constraints on the global properties of the wind.

We wish to express our gratitude to Dan Dewey and Marten van Kerkwijk for their discussions and careful reading of the manuscript and to the referee, Randall Smith, for a thorough review. J. C. acknowledges support from NASA under a GRSP fellowship. M. S.'s contribution was supported by NASA under Long-Term Space Astrophysics grant NAG5-3541. F. P. was supported under NASA contract NAS5-31429. D. A. L. acknowledges support from NASA under Long-Term Space Astrophysics grant S-92654-F. Work at LLNL was performed under the auspices of the US Department of Energy, contract W-7405-Eng-48.

REFERENCES

• Canizares, C. R., et al. 2000, ApJ, submitted First citation in article
• Decaux, V., Beiersdorfer, P., Osterheld, A., Chen, M., & Kahn, S. M. 1995, ApJ, 443, 464 First citation in article | Crossref | ADS
• Gabriel, A. H., & Jordan, C. 1969, MNRAS, 145, 241 First citation in article | Crossref | ADS
• Johnson, W. R., & Soff, G. 1985, At. Data Nucl. Data Tables, 33, 405 First citation in article | Crossref | ADS
• Kallman, T. R., & McCray, R. 1982, ApJS, 50, 263 First citation in article | Crossref | ADS
• Kawashima, K., & Kitamoto, S. 1996, PASJ, 48, L113 First citation in article | Crossref | ADS
• Kitamoto, S., Kawashima, K., Negoro, H., Miyamoto, S., White, N. E., & Nagase, F. 1994, PASJ, 46, L105 First citation in article | ADS
• Liedahl, D. A. 1999, in Proc. European Astrophysics Doctoral Network Tenth Summer School, X-Ray Spectroscopy in Astrophysics, ed. J. van Paradijs & J. A. M. Bleeker (Berlin: Springer), 189 First citation in article | ADS
• Liedahl, D. A., & Paerels, F. 1996, ApJ, 468, L33 (LP96) First citation in article | IOPscience | ADS
• Markert, T. H., Canizares, C. R., Dewey, D., McGuirk, M., Pak, C. S., & Schattenburg, M. L. 1994, Proc. SPIE, 2280, 168 First citation in article | Crossref | ADS
• Porquet, D., & Dubau, J. 2000, Rev. Mexicana Astron. Astrofis., 99, 167 First citation in article
• Pradhan, A. 1982, ApJ, 263, 477 First citation in article | Crossref | ADS
• van Kerkwijk, M. H., et al. 1992, Nature, 355, 703 First citation in article | Crossref | ADS