Astron. Astrophys. 352, 64-84 (1999)

4. Discussion

4.1. Comparison to results previously published
on ROSAT and ASCA observations

The results of the first 7.2 ks PSPC observation of NGC 4258 were published in Paper I, CWdP reported on the first 27.2 ks HRI observations of NGC 4258. The source catalogue of NGC 4258 of Paper I contains 7 of our 15 source candidates, namely X8 (source 13 of Paper I), X11 (12), X14 (14), X17 (15), X19 (16), X24 (20) and X25 (22). For all other sources but X24, the PSPC count rates of Paper I agree with the count rates given in Table 2. The source X24 is time-variable (cf. Sect. 2.5 and Fig. 4), and the lower count rate measured for X24 during the first PSPC observation block explains the difference of the count rates of this source in the source catalogue in Paper I and Table 2. With the help of the first HRI observations, CWdP detected 4 sources in the disk of NGC 4258, and their sources 3, 4, 6 and 8 correspond to our sources X24, X11, X19 and X7, respectively. Given the errors, the HRI count rates measured by CWdP agree well with our results for the first three sources, while for X7 our count rate is below the rate measured by CWdP. However, this might be due to the problem of background determination for this source which has a low signal-to-noise ratio (cf. the comment of CWdP, their Table 1). The diffuse emission components for the bulge, disk and X-ray ridge as determined in Paper I and by CWdP agree well with the results of Sects. 3.2 and 3.4.

ASCA observations of NGC 4258 with a net exposure of 36 ks were carried out in May 1993 (Makashima et al. 1994). While the good spectral capabilities and wider energy band of ASCA (0.5-10 keV) allow a decomposition of different emission components, the wider PSF function (FWHM of ) does not provide the same spatial resolution as the ROSAT data. Makashima et al. (1994) fitted a four component model to the overall NGC 4258 spectrum, (a) an absorbed power law for the nuclear component (result: ,  cm^-2,  erg s^-1 in the 2-10 keV band after removal of the absorption), (b) a thermal Bremsstrahlung component to represent the off-nucleus component of the ASCA detected hard X-ray emission ( keV,  cm^-2 (fixed),  erg s^-1), (c) a thin thermal plasma component for the soft component of the X-ray emission ( keV,  cm^-2 (fixed),  erg s^-1), and (d) a narrow Gaussian for the iron K line which is outside the ROSAT band.

In Sect. 3.2.1 the ROSAT HRI count rate for a possible point source at the position of the nucleus and surrounded by diffuse emission was calculated to  cts s^-1. From (a), one predicts a ROSAT HRI count rate of the AGN component of  cts s^-1. Assuming an absorption of  cm^-2 instead of the ASCA predicted  cm^-2, the excess ROSAT count rate at the position of the center of NGC 4258 could be explained with help of the AGN component alone. Makashima et al. (1994) attribute their emission component (b) either to very hot gas along the anomalous arms close to the nucleus or a superposition of binaries in the inner bulge region, and such a high temperature component also is indicated by the X-ray spectrum of the central region (cf. Table 7). Extrapolating the component (b) of Makashima et al. (1994) to the ROSAT band, one expects a luminosity of  erg s^-1, which is in good agreement with the ROSAT luminosity corrected for total absorption ( erg s^-1 for the PSPC measurement,  erg s^-1 for the HRI). The component (c) of the ASCA data should contribute a luminosity of  erg s^-1 in the ROSAT band, which again is in agreement with the bulge luminosity measured by ROSAT (cf. Table 9) after subtracting the central emission region.

4.2. Superluminous sources in NGC 4258

The maximum X-ray luminosity that can be achieved in an accreting binary system, assuming steady spherical accretion, was calculated by Eddington (1928) as  erg s, where M is the mass of the accreting object. Assuming a maximum mass of for a neutron star, the Eddington limit is  erg s^-1. Three point-like sources in NGC 4258, X8, X16 and X24, exceed this limit, thus they are `superluminous'. The time variable sources X16 and X24 have maximum luminosities of  erg s^-1 and  erg s^-1, respectively. While the time variability strengthens the binary hypothesis in the case of X16 and X24, variability for X8 could neither be established from the individual HRI or PSPC blocks nor by a comparison of the mean HRI and PSPC fluxes. If X8 were no X-ray binary, the source could be a supernova remnant expanding into a high density medium with  cm^-3 (cf., e.g., Vogler et al. 1997 for the expected luminosities of such a SN remnant).

The X-ray spectra suggest either a thermal Bremsstrahlung or a power law spectrum for X24 (cf. Table 8). The unabsorbed maximum luminosities are  erg s^-1 and  erg s^-1 for the thermal Bremsstrahlung and power law model, respectively. The predicted masses of the central object would be in excess of . We searched for optical counterparts of the X-ray source on a deep optical plate kindly provided by Arp (1994). An overlay of the PSPC contours calculated from the initial 7.2 ks PSPC observation was shown in Fig. 2 of Paper I. The deep PSPC observations confirm the attitude solution suggested in that figure, and the PSPC emission maximum is centered on a pattern of knots visible in the outermost spiral arms of the galaxy. The HRI suggests a slightly different attitude solution for X24, and the source then would be located at the border of the knots. If the structures visible in the deep optical image are at the distance of NGC 4258, and not due to fore- or background objects, they may be interpreted in terms of HII regions. Star formation in such a region indeed is expected to be responsible for the existence of superluminous sources like X-ray binaries or supernova remnants expanding into high density media. Other cases of extremely bright sources correlated to giant HII regions were also reported for NGC 4559 (Vogler et al. 1997), NGC 4631 (Wang et al. 1995, Vogler & Pietsch 1996) and M101 (Wang et al. 1999). While no time variability could be established for the X-ray source in NGC 4559, the sources in NGC 4631 and M101 are variable as is the case for NGC 4258. However, contrary to the case of NGC 4631, spectra of the optical candidates are still needed for the other galaxies to exclude the possibility of background objects shining through the outer spiral arms. If then the time variable sources are indeed proven to be correlated to HII regions, these superluminous binaries would be another `species' of extremely bright X-ray sources besides active galactic nuclei and supernovae/SN remnants. Extremely bright supernovae ( erg s^-1) were reported for NGC 891 (Bregman & Pildis 1994), NGC 1313 (Colbert et al. 1995), NGC 6946 (Schlegel 1994) and M81 (Zimmermann et al. 1994)

4.3. Comparison of the NGC 4258 point source population
to other galaxies

The local group spiral galaxies M31 (Supper et al. 1997,  Mpc) and M33 (Schulman & Bregman 1995,  Mpc) and the very nearby galaxy NGC 253 (Vogler & Pietsch 1999,  Mpc) were investigated with ROSAT to lower X-ray point source luminosities than NGC 4258. The PSPC data were used to calculate luminosity distributions of M31 and M33. For NGC 253 as well as NGC 4258 the associated PSPC and HRI observations reach approximately the same detection limit; thus we used the HRI data providing better spatial resolution. While 340 (26, 25) sources above a detection limit of  erg s^-1 ( erg s^-1,  erg s^-1) are contained in the luminosity distributions of M31 (M33, NGC 253), only 12 HRI point sources are contained in the NGC 4258 distribution (detection limit  erg s^-1). We compare the NGC 4258 results also to the case of other, more distant spiral galaxies which have been part of a survey program for hot interstellar medium and point sources performed at the Max-Planck-Institut für Extraterrestrische Physik (e.g. Vogler 1997, Immler 1999). The individual galaxies, their parameters like distance, morphological type, optical extent or blue luminosity as well as the integrated X-ray point source luminosities, after excluding the bulge regions and fore- or background objects, are compared to M31, M33, NGC253 and NGC 4258 in Table 10. References of ROSAT X-ray papers for the individual galaxies are included in the table. From Einstein results, Fabbiano et al. (1983) established proportional to . However, for individual galaxies a strong scatter in the correlation is seen. Most of the galaxies presented in Table 10 have ratios between and . Since nuclear X-ray emission was excluded from the point source luminosity, the active galaxies with bright optical nuclei, namely the nuclear star burster NGC 253, the Seyfert I galaxy NGC 1566 and the Seyfert II galaxy NGC 4258, tend to populate the lower range of the measured ratios. With ratios above , NGC 4559 and M100 have the highest X-ray point source to blue luminosity ratios. In the case of NGC 4559, the high ratio can be explained by a superluminous source in the outer spiral arms ( erg s^-1), and two extremely bright sources with luminosities in excess of  erg s^-1 are also is detected in M100.

Table 10. Non-nuclear X-ray point source conributions in nearby spiral galaxies.
Notes:
) Tully (1988)
) Dickey & Lockman (1990)
) 0.1-2.4 keV luminosity assuming a thermal Bremsstarhlung spectrum ( keV), corrected for Galactic foreground absorption
) References: a) Supper et al. (1997), b) Schulman & Bregman (1995), Long et al. (1996), c) Ehle et al. (1995), d) Immler et al. (1999), e) Immler et al. (1998), f) Wang et al. (1999), g) Vogler & Pietsch (1999), h) Ehle et al. (1996), i) This work, j) Vogler et al. (1997), k) Vogler et al. (1996), l) Vogler & Pietsch (1996)

To search for possible differences in the point source population of the investigated galaxies, a luminosity distribution diagram was calculated (Fig. 8). To take the different sizes of the galaxies into account, the number of the sources per energy interval was normalized to the blue luminosity of the galaxies, which, as mentioned, is expected to be proportional to the total X-ray luminosity. The luminosity distributions of all galaxies are comparable with the exception of the lowest luminosity bin of M33. This bin, however, might be affected by the relatively low detection limit and different source detection technique used for this galaxy by Long et al. (1996).

Fig. 8. Luminosity distributions of detected point sources (excluding nuclear sources) for the spiral galaxies presented in Table 10 (cf. this table for references). The error bars (1) for the individual points were calculated from either a Gaussian error distribution or Poisson statistics when appropriate (cf. Gehrels 1986)

The brightest point source in NGC 4258 is more luminous than the individual sources in M31, M33 or NGC 253. On the other hand, as is the case for the latter galaxies, we detect no extremely luminous ( erg s^-1), non-nuclear point source in NGC 4258 in opposite to NGC 4559, NGC 4565, M83, and M101. Assuming that the sources with  erg s^-1 are really point like, as ROSAT HRI observations suggest, the absence of such sources in several galaxies can either be explained in terms of time variability or by a differing source population.

Contrary to the timing investigations performed for NGC 253, which proved time variability for roughly one third of the NGC 253 sources, the results for NGC 4258 allow us to establish time variability for only three sources. In the case of NGC 253 nearly all sources with a luminosity above  erg s^-1 turned out to be time variable, and from their light curves these variable sources were classified as X-ray binaries. However, the low detection rate of time variability in the case of NGC 4258 is masked by the clearly worse photon statistics for sources of similar luminosity compared to NGC 253. The expected HRI count rate for a source with  erg s^-1 is  cts s^-1 (70 cts in 50 ks) for NGC 253, while in the case of NGC 4258 the expected count rate of  cts s^-1 would only yield 13 counts during the 54.8 ks HRI observation; thus preventing the detection of time variability.

4.4. Non-detection of the type II SN 1981 K

The supernova (SN) 1981 K was first detected at radio wavelengths (van der Hulst 1983). It was classified as a type II supernova, and radio light curves were presented in van Dyck et al. (1992) and Calle et al. (1996). The source is located 18:002 east and 75:002 north of the nucleus of NGC 4258, and its position is marked with a cross in Figs. 2. No X-ray source at the position of the SN, which is surrounded by enhanced diffuse X-ray emission (cf. Figs. 1-3), was found by the detection algorithm. To calculate a 2 upper limit for a possible X-ray source at the position of SN 1981 K, the HRI counts in a detect cell of diameter were compared to a local background from to diameter, yielding an upper limit (2), corrected for deadtime, vignetting and exposure, of ( erg s^-1).

4.5. Comparison of the X-ray point source list
to other wavelengths

We compared our ROSAT point source catalogue to radio (Willis et al. 1976) and H (Courtès et al. 1993) maps of NGC 4258. Possible counterparts for X11 (radio source 1216+47W2) and X19 (HII region no. 54) already were given in Paper I. HII region no. 6 is proposed as only additional counterpart for the new source X16 ().

4.6. The anomalous spiral arms
and a possible halo component

A close correlation of the diffuse X-ray emission to the anomalous spiral arms was reported from the first short 7.2 ks ROSAT PSPC observations (Paper I) and the first 27.2 ks HRI observation (CWdP). Paper I attributed the X-ray emission partly to hot gas along the anomalous spiral arms and to hot gas emerging from the plane of the galaxy into the eastern halo of the galaxy, the latter component being mainly visible in ROSAT PSPC S band. CWdP attribute the main part of the diffuse emission to the anomalous spiral arms. In their model the diffuse X-ray emission is caused by shock heating of the interstellar medium via jets emerging from the active nucleus, and the absence of X-ray emission west of the north western tip of the arm is explained by a lower dissipation of the jet in this region. If the anomalous arms are not caused by jets but by bar shocks, as Cox & Downes (1996) suggest, the heating process of the interstellar medium could either be (a) the bar shock itself or (b) regions of interstellar medium in which star formation (and at a later stage supernova explosions) is triggered by the bar shocks. Cox & Downes (1996) argue that, due to the Maxwell distribution of the shock velocities, a small fraction of high velocity shocks accounts for heating parts of the interstellar to million K temperatures, and therefore they favor the explanation (a).

The contours of a 1480 MHz radio map of the NGC 4258 region (Albada & van der Hulst 1982) overlaid on the ROSAT HRI image smoothed with a Gaussian filter of FWHM (cf. Fig. 2) is shown in Fig. 9. Due to systematical errors in the X-ray source positions and the radio positions, residual errors of the relative X-ray to radio positions in Fig. 9 cannot be excluded. However, the close correspondence of the X-ray and radio emission peak at the position of the nucleus of NGC 4258 and the detection of a radio double peak in the X-rays (source X11) confirms the relative pointing solution. This solution is also supported by the coincidence of the radio peak south east of X19 with an enhancement of X-ray emission visible in Fig. 2 (the position of this enhancement is marked with a cross in Fig. 9). The significance of the X-ray enhancement,  cts s^-1, however, is too low to classify this emission as a point source, for which we required a Gaussian significance (, cf. Sect. 2.4.1). The correlation of the X-ray emission to the anomalous arms is evident, and the bifurcation of the anomalous arms known from radio and H observations (marked with an arrow in Fig. 9) seems to be traced also in the X-rays.

Fig. 9. Contours of an 1480 MHz radio map of the NGC 4258 region (Albada & van der Hulst 1982) overlaid on the ROSAT HRI image presented in Fig. 2. Point sources are marked according to Table 2 and Fig. 1

The X-ray data do not allow to finally decide if the observed hot gas along the anomalous arms is caused by (a) bar shocks, (b) enhanced star-formation triggered by the bar shocks, (c) dissipated energy of a jet or (d) a hyperbubble of hot gas, the emission of which could be partly shadowed by the HI disk. Following the argumention (b), Cox and Downes (1996) conclude that the real bar has an extent of along the anomalous arms, and they interpreted the bend-off of the anomalous arms, also visible in the X-rays (especially to the south east), as projection effects (inclination of the galaxy 72^o) due to gas flowing out of the plane of the galaxy in the z direction. In the picture of Cox & Downes (1996) the steep radio gradients are the leading edges of the bar shock. These gradients, on the other hand, are the northern (southern) boundary of the north western (south eastern) X-ray ridge. Not only the steep leading edges of the bar shocks emit X-rays, but also the regions behind this shock fronts. This might indicate that both, hot interstellar medium due to the bar shocks, as well as hot interstellar medium created in regions of enhanced star formation triggered by the bar shocks are visible in the X-ray light. A comparison of the diffuse emission of NGC 4258 with NGC 253 might shed additional light on the interpretation of the anomalous arms in terms of a nuclear hyperbubble (d). In the edge-on starburst galaxy NGC 253 the outflow from the nuclear region is visible as a more symmetric structure. Assuming that the absorption by cold gas should on average be lower for the inclined galaxy NGC 4258, it remains difficult to explain the asymmetry in the case of NGC 4258. However, as Sofue (1984, 1999) points out in the case of the Milky Way, a part of the assymetry might be caused by the rotation of the galaxy.

Besides CO and radio observations, H observations allow us to trace the anomalous spiral arms. An H emission map (Dutil & Roy 1999) was overlaid with the HRI X-ray contours (Fig. 10), and a close correlation of the X-ray and the H emission is as evident as in the case of the X-ray and radio emission.

Fig. 10. ROSAT HRI contours overlayed on an H map of NGC 4258, kindly provided by Dutil & Roy (1999). The contours were calculated from the image presented in Fig. 2 and represent levels of 2, 3, 5, 7 and  cts s-1 arcmin- 2 above the background. Point sources are marked according to Table 2 and Fig. 1, the position of the nucleus (Table 1 of Paper I) of the galaxy with a cross

For a cylindrical symmetry enclosing the X-ray ridge at distances of from the nucleus for a radius of , the measured X-ray luminosities and temperatures can be translated into a density and mass of the gaseous component. Assuming a spherical distribution of the gas and radiation equilibrium (which, however, might be an oversimplification due to the shock heating of the interstellar medium), we find  cm^-3 / and  M, where the filling factor allows for clumpiness of the gas in the volume filled by hot gas (cf. Nulsen et al. 1984 for the used formulae). The values are in good agreement with values received by CWdP for a slightly different geometry.

The ROSAT PSPC S band (cf. Fig. 6), showing diffuse emission with an emission maximum centered east of the nucleus, can be interpreted in light of a ballistic model of the jets. If hot gas is escaping the galactic plane at the south eastern edge of the X-ray ridge, and, as the absorption suggests, this gas is located above the disk of the galaxy, the PSPC S band might trace both, diffuse X-ray emission from gas along the anomalous arms and hot interstellar medium in the outer disk/halo of the galaxy. Then, the offset of the S band emission maximum from the center of the galaxy along the minor axis can be explained by halo gas filling the front hemisphere of the inclined galaxy. The offset of the maximum along the major axis might be explained by the fact that the gas escapes near the south eastern tip of the X-ray ridge, and not uniformly out of the entire disk of NGC 4258 (cf. also Paper I).

Assuming that at least one half of the soft component of the two component spectrum of the diffuse emission (cf. Tables 6 and 7) is caused by a halo component distributed in a hemisphere with a radius of , one calculates  cm,  M, and  yr for the  K gas in the halo of NGC 4258 (again following Nulsen et al. 1984 in converting the measured luminosity and temperature to mass, density and cooling time).

One might expect that, similar to the front hemisphere, the reverse hemisphere of the NGC 4258 is filled with hot gas escaping near regions of the north western tip of the anomalous arms. However, assuming a temperature as estimated for the front hemisphere, such a component would not be detectable since the absorption of these very soft X-rays by cold gas in the disk would be very high (the absorption cross section is proportional to for photons of energy E in the ROSAT band).

4.7. A combination of the spatial resolution of the HRI
with the spectral resolution of the PSPC
for the diffuse emission features

To visualize the ROSAT HRI and PSPC results obtained for the anomalous spiral arms in one image, a `true color image' of the diffuse X-ray emission along the arms was calculated (cf. Fig. A1 (Plate 1)). We tried to combine the HRI (spatial) and PSPC (spectral) results. In a first step, the background subtracted HRI image was multiplied with the background subtracted PSPC soft, hard1 and hard2 image, the resulting images will be referred to as HRI-S, HRI-H1 and HRI-H2. Each image was subdivided in 6 intensity levels, and a color was attributed to the subbimages: Red for HRI-S, green for HRI-H1 and blue for HRI-H2. To calculate a true color image from the three images, a color described by the six red, green and blue levels was calculated for each pixel of the result image. Therefore, the color image has 216 possible colors, from black (no emission) to white (maximum emission in all energy bands). Sources with emission maxima in the PSPC soft (hard1, hard2) band shine up in reddish (greenish, bluish) color etc.

The true color image demonstrates the spectral hardening of the X-ray emission from the south east to the north west tip of the X-ray ridge. One explanation for this spectral hardening is the superposition of emission from the anomalous arms and the relatively smooth soft band emission at least partly reflecting a halo component of the X-ray emission and being mainly visible in the south eastern halo hemisphere. However, given the small spatial scale of the variations of the spectral hardness in comparison to the smooth soft band morphology, the spectral hardening might be due to changes in the gas temperature or the absorption of the X-ray emission along the anomalous arms. Assuming a relatively constant gas temperature along the arms, the spectral hardening would be a measure for increasing absorption along the arms from the south eastern to north western side. Such a finding might support the ballistic interpretation of the anomalous arms, which assumes that these are located above (below) the disk of NGC 4258 at the south eastern (north western) tip of the arms.

We tried to quantify the findings suggested by the true color image and extracted 7 spectra along the anomalous spiral arms. As expected, the values of one component models (THBR, THPL, POWL) were in general not acceptable. However, we failed in introducing two component models (namely the superposition of a THPL component describing the halo emission and either a THPL or THBR component of higher temperature describing the anomalous arms): The errors in the individual spectral parameters were too high to allow final conclusions, and this reflects the limited spatial and spectral resolution of the PSPC data.

4.8. Comparison of the diffuse emission components
of NGC 4258 to those of other spiral galaxies

The diffuse X-ray emission along the anomalous spiral arms covering the bulge and inner spiral arms of NGC 4258 can be compared to diffuse X-ray components detected in the bulges of other spiral galaxies. Table 11 gives the reference publications and derived parameters of other galaxies in comparison to NGC 4258. The nearly edge-on starburst galaxy NGC 253 hosts a giant nuclear superbubble and plumes of diffuse X-ray emission protruding from the bulge into the disk and/or halo of the galaxy, most probably being connected to hot gas flowing out from the central star formation region. While the anomalous arms of NGC 4258 have a position angle close to that of the major axis of the galaxy, the plumes of NGC 253 are extended perpendicular to the major axis of the galaxy. Diffuse emission out of the bulge of the starburst galaxy NGC 1068 (inclination ) was reported by Wilson (1992) ¹, and the X-ray emission of the bulge has been interpreted as the superposition of a point-like contribution of the the Seyfert I nucleus of NGC 1068 and diffuse X-ray emission caused by the central starburst disk. Similar to NGC 253, starburst activities in NGC 2146 () are responsible for diffuse X-ray emission in the bulge of the galaxy, and the X-ray emission features in the latter galaxy are proposed to have the same origin as the plumes in NGC 253. In the case of the edge-on starburst galaxy NGC 3079, a superbubble of hot gas similar to NGC 253 is reported. The bubble in NGC 3079 is displaced slightly from the center of the galaxy, and the Seyfert II /LINER nucleus possibly adds minor contributions to the detected diffuse ROSAT HRI emission. For the purpose of Table 11, the entire diffuse HRI emission has been attributed to the superbubble. Two further examples of diffuse X-ray emission in the bulges of spiral galaxies most probably caused by a hot thermal component are the galaxies M51 () and M83(). M51 has a weak Seyfert II nucleus similar to NGC 4258, and M83 is a galaxy with enhanced star formation.

Table 11. Hot interstellar medium in the bulges/inner disks of spiral galaxies.
Notes:
) Luminosities (0.1-2.4 keV) corrected for Galactic foreground absorption
) References: a) Vogler & Pietsch (1999), Pietsch et al. (1999), density and mass from Vogler (1997), b) Wilson et al. (1992), c) Armus et al. (1995), d) Pietsch et al. (1998), density and mass from Vogler (1997), e) This work, f) Ehle et al. (1995), g) Ehle et al. (1997)
) ROSAT HRI measurment within the contours of HRI image of Fig. 2
) Publications based on HRI data allone. Due to the lack of spectral capabilities, these data do not enable the measurement of a temperature
) The geometry of the emission regions is very difficult to judge, no values given by the authors
) No values given by the authors

The surface brightness of the diffuse emission along the anomalous spiral arms of NGC 4258 ( erg s^-1 kpc^-2) is similar to that along the plumes of NGC 253, along the extended emission in NGC 2146 and the diffuse X-ray emission reported by Wilson et al. (1992) from the central starburst disk of NGC 1068. The surface brightness seems to be higher for the nuclear superbubbles of NGC 253 and NGC 3079 ( erg s^{- 1} kpc^-2), as well as in the case of the diffuse bulge emission in M51 and M83.

Due to the lower inclinations, the emission components and their geometry is very difficult to judge in M51 and M83, NGC 1068 and NGC 2146. The temperature of the X-ray components for the inner bulge regions are  keV for all galaxies. In the case of NGC 253, NGC 3079 and NGC 4258 the geometrical interpretation of the measured emission components is relatively simple, thus allowing the determination of the electron density and mass of the thermal gas. For NGC 3079 and NGC 4258, those densities are around  cm^-3, and the predicted density is slightly higher in the case of central superbubble of NGC 253 ( cm^-3). Taking the possible uncertainties in the geometries as well as in the filling factors into account, no principal difference can be seen for the three galaxies. However, the masses and luminosities of the gas contained in the superbubble of NGC 3079 and along the anomalous spiral arms of NGC 4258 are about one order of magnitude higher than the values measured for the central superbubble of NGC 253. While there is no direct evidence for such a contribution, this might indicate the importance of the nuclear activities in NGC 3079 and NGC 4258, possibly contributing to the power input into the hot interstellar medium.

The second major diffuse emission component besides the X-ray emission along the anomalous spiral arms is the halo component of NGC 4258 indicated by the PSPC S band for the eastern halo hemisphere. The estimated density ( cm^-3) and temperature ( keV) of the halo gas are comparable to values measured for other spiral galaxies like NGC 253, NGC 891, NGC 3079, NGC 3628, NGC 4449, NGC 4631, M51 and M83 ( from  cm^-3 to  cm^-3; T from 0.1 keV to  keV; overviews are given, e.g., in Vogler 1997, Dahlem 1997). From this, there is no evidence for a different physical state of the halo gas depending on the different `feeding' mechanisms (the anomalous arms in NGC 4258 versus star formation/superbubbles in the other galaxies).

4.9. Shadowing of the soft X-ray background
by the outer disk of NGC 4258

In the outer disk of NGC 4258 we measure a depression of  cts s^-1 in the PSPC S band (cf. Sect. 3.4). Given the area of the disk after exclusion of the point sources (31.7 arcmin²), this translates to a negative surface brightness of  cts s^{- 1} arcmin^-2.

The HI disk of NGC 4258 has an estimated surface density of  cm^-2 in the region of the outer disk (van Albada 1980b), thus the transmission of X-ray photons in the 0.1-0.4 keV range is below 7%. This should lead to a strong depression of the X-ray background, which originates from sources behind the disk of NGC 4258, in the soft band, but not in the hard band, as observed.

A similar effect was found by Snowden et al. (1994) for a region in Ursa Major, by Supper (1995) in M31, by Warwick (1995) in NGC 55, by Snowden & Pietsch (1995) in M101 and by Vogler et al. (1997) in NGC 4559. They estimate the surface brightness of the soft X-ray background as  cts s^-1 arcmin^{- 2},  cts s^{- 1} arcmin^-2,  cts s^-1 arcmin^{- 2},  cts s^-1 arcmin^{- 2}, and  cts s^-1 arcmin^{- 2}, respectively. Their values were corrected for Galactic HI absorption and assume a power law spectrum with an index of -1.96. For this model one estimates the background brightness deficiency of the outer disk of NGC 4258 to  cts s^-1 arcmin^{- 2}, and this value should be perceived as a lower limit for the soft X-ray background since the measured X-ray brightness of the outer disk may still contain residual emission from the galaxy, and because the diffuse X-ray background may not be completely shielded by the HI disk. In any case, our estimate is in good agreement with the measurements of others.

© European Southern Observatory (ESO) 1999

Online publication: November 23, 1999
helpdesk.link@springer.de