3 Results

The results of our observations are shown in Table 1, which lists some designations of the studied H II regions, the equatorial coordinates of the centres of the observed areas, the diameter of the circular diaphragm used, the number N of repetition of the basic observing routine (or the integration time per filter in units of 160 s), the equivalent width $W_{\rm H\beta}$, the [O III]/H$\beta$ line ratio and the logarithmic H$\beta$ flux in units of erg s^-1 cm^-2. The error estimates of these properties correspond to the propagation of the (1$\sigma$) statistical errors of the counts in each filter. The sky subtraction is the most important source of uncertainty, specially for the equivalent width $W_{\rm H\beta}$, due to the inevitable inclusion in the large apertures used of relatively bright field stars. In order to minimise this problem we have taken different sky positions around each nebula. In some cases the measurements of $W_{\rm H\beta}$ were ruin by severe brightness fluctuations of the background.

 

 

Table 2: Lyman continuum photon fluxes

Object	D (kpc)	C(H$\beta$)	$\log$ $N_{\rm c}$ (s-1)
NGC2467	4.5 [1]	0.7 [6,14,15]	48.9
RCW19	3.0 [1,2,3]	1.3 [3,6]	47.8
NGC2579	3.3 [2,3,4]	1.1 [3]	47.8
RCW34	2.9 [5]	1.9 [5,6]	47.8
Gum22	1.7 [2]	3.0 [6,16]	48.4
RCW39	3.0 [6]	3.2 [6]	48.9
RCW40	1.7 [1,2]	1.4 [6]	48.2
NGC3199	3.5 [2,7]	1.3 [17,18]	48.9
NGC3503	2.8 [8]	0.7 [8]	46.6
NGC3603	7.7 [2]	3.6 [19]	51.3
RCW104	3.7 [2]	1.8 [17]	47.9
Gum64a	2.0 [2,9,10]	1.7 [9,10]	47.3
Gum64c	2.0 [2,9,10]	1.7 [9,10]	47.6
NGC6357	1.7 [2,10]	2.5 [16]	48.6
M20	1.8 [11,12,13]	0.9 [15,20]	48.7

References: [1] Brand & Blitz (1993); [2] Avedisova & Palous (1989); [3] Moffat et al. (1979); [4] Neckel & Staude (1995); [5] Heydari-Malayeri (1988); [6] Shaver et al. (1983); [7] Hidayet et al. (1982); [8] Herbst (1975); [9] Walborn (1982); [10] Neckel (1978); [11] Buscomb (1963); [12] Lada & Wooden (1979); [13] Walborn (1973); [14] Peimbert et al. (1978); [15] Hawley (1978); [16] Danziger (1974); [17] Esteban et al. (1992); [18] Kwitter (1984); [19] Girardi et al. (1997); [20] Lynds et al. (1985); Note: E(B-V) converted to C(H$\beta$) by C(H$\beta$) = 1.5 E(B-V).

Table 2 presents our estimates for the Lyman continuum photon fluxes $N_{\rm c}$ calculated from our H$\beta$ line fluxes for those objects with the additional data required available in the literature, namely the heliocentric distance, D, and the logarithmic extinction in the H$\beta$ emission line, C(H$\beta$). We must stress that all these objects show extended emission outside the observed area. So, our values should be more properly considered as lower limits for $N_{\rm c}$. Nearly five orders of magnitude separate the ionising powers of NGC3503 and NGC3603, one of the most luminous H II region in the Galaxy excited by more than 50 O stars. The objects in the sample have also shown varied degrees of excitation even at similar galactocentric distances, with [O III]/H$\beta$ ranging from 0.3 to 10. The large majority has shown $W_{\rm H\beta} \leq$ 100 Å, which according to the models of Copetti et al. (1986) with a normal IMF slope indicates that they are evolved objects with ages around or larger than 4$\,\,10^6$ years.

In Fig. 1 we have shown that our [O III]/H$\beta$ measurements for both the programme H II regions and the planetary nebulae of the controlling sample are very well correlated with the mean spectroscopic values found in the literature with a correlation coefficient R = 0.99. Moreover, the regression line, [O III]/H$\beta$ (other authors) = $(0.03\pm0.41) + (1.02\pm0.04)$ [O III]/H$\beta$ (this paper), is statistically indistinguishable from the identity line, which justifies the usual procedure of adopting for the whole nebula data collected from a small region. Of course, every line measurement is a weighted integration along the line of sight. In particular, the similarity between the integrated and non-integrated [O III]/H$\beta$ ratios may in great part be attributed to the fact that in normal H II regions the O⁺⁺ zones usually occupy large fractions of the total nebular volumes. We would expect to find more discrepant comparison of this sort among measurements of emission lines produced by more localised ions (e.g., [O I] and [O IV] lines).

Acknowledgements

This work was partially supported by the Brazilian institutions CNPq and FAPERGS. We thank the referee for very helpful comments and suggestions.