5 Types of selected objects

Let us investigate in more detail the different types of ${\rm H}\alpha$ objects. In Table 2 we present their mean photometric parameters. The columns give morphological types of objects, their numbers, stellar magnitude V, colour indices U-B, B-V (IFM) and V- ${\rm H}\alpha$, surface brightness ( ${\rm H}\alpha$ flux per square arcsecond), ${\rm H}\alpha$ excess parameter S, size of objects and ${W'_{\rm H\alpha}}$. Stellar magnitude in ${\rm H}\alpha$ is not calibrated and it has been derived from the relative flux in this band, m(H$\alpha$) = -2.5 lg F(H$\alpha$)+24.5. It should be borne in mind that the flux F( ${\rm H}\alpha$) depends directly for extended objects on a nebula size, but this is not the case for stars as point sources.

As distinct from the continuous spectrum of stars, where there is a direct relationship between colour indices and temperature of a star, the colours of a nebula depend on a number of factors including the ionizing star temperature, electron density, dilution factor (size of a nebula), optical depth. Besides, colours may vary in a complex manner depending on the reddening. The intensity of hydrogen Balmer lines does not depend strongly on the nebula temperature. Together with hydrogen lines, the forbidden lines [OIII] $\lambda$ 4959, 5007 are the most intensive ones in the spectra of nebulae and a good indicator of electron temperature of gas (Allen 1973). They are approximately equal contributors to B and V bands. Other strong lines do not fall within V band. In B band are the intensive lines of a hot gas: H$\beta$, [ArIV] $\lambda$4712, [NeIV] $\lambda$4725, HeII $\lambda$4686, [OIII] $\lambda$4363 and [NeIII] $\lambda$3967, 3969. The last two lines make about the same contribution to U band. There is the line [OII] $\lambda$3727 in U band. With growing electron temperature and excitation of the nebula gas, the flux in B band will dominate. Accordingly the value of U-B will increase while B-V will decrease.

Figure 4: Two-colour diagram for stars (full circles), from the bottom upwards for bright, mean and weak groups, respectively; diffuse nebulae (circle), bubbles (square), common intermediate objects (rhomb) and zs type objects (filled rhomb). Consequences of I and V luminosity classes and their reddening lines are shown

It is seen from Table 2 that in U-B and B-V the diffuse objects differ greatly from the bubbles. This is also displayed in Fig. 4. From the colours the diffuse nebulae are cool, their behaviour in the figure resembles the behaviour of the stars but the diffuses have a lower value of U-B. This may be due to the emission Balmer jump and the strong hydrogen lines in their spectra. The latter is suggested also by the large value of S in the diffuse nebula (Table 2).

In contrast to the diffuse objects, colour indices of bubbles indicate that the gas excitation degree in them is high. Besides, the objects of type b are, on average, considerably fainter in V than the diffuse nebulae. The location of these objects on the diagram in Fig. 4 suggests that they are clearly different in excitation and that the central star is a minor contributor to the spectrum. If the heating of such nebulae is radiative, the stars having a relatively small size must be sufficiently hot. It is unlikely, however, that these b nebulae sizes are defined at all by the central star's temperature and luminosity (see below Fig.9). In the gas excitation collisional processes have to be dominating. From their specific ${\rm H}\alpha$ morphology (bubbles), colour indices, low luminosities, as well as their sizes and surface brightness behaviour (see below) we can refer them to envelopes around WR stars and SN remnants.

No less than 80$\%$ of WR stars in the Galaxy are embedded in HII regions (Lozinskaya 1992), and no less than 40$\%$ having ring nebulae. If the bubble-type nebulae surround WR and Of stars, then in heating and excitation mechanisms they may be similar to planetary nebulae. In this case we propose that a star itself could be not seen in the visible region of the spectrum and the observed emission is produced by the nebulae. In the case of extremely hot central stars of planetary nebulae, the star is fainter than the nebulae by ${ \bigtriangleup B=6^m}$, and at the star temperature of about $3\ 10^4$ K the star is fainter by ${\bigtriangleup B=0\hbox{$.\!\!^{\rm m}$ }4}$ (Allen 1973). It can also be concluded that on U, B and V plates these objects must look like diffuse ones. That is why the largest of these objects are likely to be missing in the catalogue (IFM). It may be for the same reason the b nebulae have smaller sizes than d ones.

In the group c (faint and compact objects) all types of objects are probably present, however, compact diffuse nebulae are apt to be the most numerous among them. It is very remarkable that the zs-type objects that have been classified as a separate group only by ${\rm H}\alpha$ image morphology have the bluest index U-B, the reddest index B-V and they are the most powerful ${\rm H}\alpha$ sources (see Table 2). Maybe these are compact HII regions with a very high gas density, with hydrogen emission lines of a nebula being a powerful contributor to the spectrum.

Figure 5: Standard indices as a function of V - H$\alpha$. Here, as in the following figures, stars are marked with dots and solid lines, diffuse nebulae - rhombs and dotted lines. The b nebulae (long-dashed lines) are not presented

Figure 5 shows the relationship between the standard colours indices and V- ${\rm H}\alpha$. In this figure and those followings the stars are designated by dots and their approximations by solid lines, the diffuse nebulae by rhombs and dotted lines, the b nebulae by squares (if shown) and longdashed lines. Here and further in the analysis of such relationships we use the linear function ${ y=C_x\,+\,C_1}$ and present only the value of C and its rms deviation. The stars in Fig. 5 behave as one may expect: with increasing blue excess the intensity of ${\rm H}\alpha$ grows. For the stars ${ C(U\!-\!B, \,V\!-\!{\rm H}\alpha) = -0.085\pm0.035}$ and ${ C(B\!-\!V, \,V\!-\!{\rm H}\alpha)}$ ${ =-0.083\pm0.029}$. These relationships for the nebulae of b and d types differ considerably from those of stars, which confirms that these are the objects of different nature. Only in the diffuse nebulae a significant dependence is observed, ${ C(U\!-\!B, \,V\!-\!{\rm H}\alpha) = -0.109\pm0.019}$, which is likely to be due to the growing size of the HII region with increasing temperature of the exciting star.

In Fig. 6 a distribution of the selected objects according to the galactocentric distance ${ R_{\rm gc}}$ is displayed, that is the number of objects within the distances intervals. The corresponding distributions of the object numbers per unit area would be similar, but it would have quite a strong peak at small ${ R_{\rm gc}}$, which would make it difficult to analyze. The object coordinates were deprojected from the picture plane onto the galactic plane. It was adopted according to Vaucouleurs (1959) that the inclination angle of the galactic rotational axis to the line of sight $i=55\hbox{$^\circ$ }$ and the position angle of the major axis is PA  $= 23\hbox{$^\circ$ }$. It is important to note here that in the central regions of the galaxy the real density must be higher. The selection effect is due to identification problems against the strong ${\rm H}\alpha$ background of the central parts (the first 1-2 bins in Fig. 6). This selection is about the same for all types of objects, however it affects mainly the diffuses and bubbles, whose average surface brightness is not high. It follows from the figure that the three types of objects belong to different populations. The distribution of emission stars along the radius is relatively uniform. The observed minima in the distribution (if even they are real) are insignificant in our data and could be random. The density of stars falls by a factor of 2 at ${ R_{1/2}\approx4.2}$ kpc ( $1{}^{\prime\prime}$ = 3.5 pc). The diffuse nebulae distribution is quite different, they are located, on average, closer to the centre, ${ R_{1/2}\approx2.4}$ kpc, and they show a single significant maximum at 1.8 kpc.

Figure 6: Distribution of the objects in intervals of galactocentric distance, solid line -- stars, dotted line -- diffuse nebulae, long-dashed line -- b type nebulae. The upper axis is graduated in kpc

The objects of type c are not presented in Fig.6, they resemble most of all the diffuse nebulae. The nebulae of type b exhibit the most broad distribution with ${ R_{1/2}\approx4.7}$ kpc. It has two maxima at 2.9 and 4.5 kpc. The maxima are formally insignificant, however, the bubbles distribution itself is considerably more irregular than that of the diffuse ones (with about the same number of objects). It is not improbable that these maxima display that predominantly spiral arms fall within the defined intervals of galactocentric distances, or possibly they are due to ring structures in the galaxy. Another remarkable detail in Fig. 6 is that maxima in the distribution of stars and diffuse nebulae fall within the regions 1.4-2.1 and 3.5-4.2 kpc ( $400{}^{\prime\prime}-600{}^{\prime\prime}$ and $1000{}^{\prime\prime}-1200{}^{\prime\prime}$, respectively) and coincide with the minima in the distribution of bubbles. This matter is discussed below.

Thus, a number of characteristics: colour indices, surface brightness, distribution in the galaxy suggest that s, b and d objects are physically different. For instance, in ${ V-\rm H{\alpha}}$ the b and d nebulae differ by $8\sigma$ (Table 2). Location of these objects over the galaxy is not uniform.