Figure 10 shows the relation between the logarithmic HH luminosity of a given HII
region, log L (with L in erg s-1), and the logarithm
of the cube of the radius log r3 (r in arcsec). If this relation were to
maintain a constant slope, this would mean that the product of the electron
density, the proton density and the filling factor (fraction of the volume
filled with ionized medium, Osterbrock 1974) were constant, independent of the
luminosity of
the HII regions. If regions of high luminosity were density bounded
it would yield an increasing slope with increasing
volume, i.e. the relation should show a concave upward curve, corresponding to an increase in the
average filling factor. Although the
scatter in Fig. 10 is large, there is an upward curvature at high
luminosity, which is in agreement with the general scenario in which the most
luminous HII regions in spirals (and in all probability in irregulars, see Fuentes 1997) are
density bounded (Beckman et al. 1999; Rozas et al. 1996a). The
observed distribution is discretized along the abscissa in Fig. 10 because r
is calculated by counting the number of pixels within a region,
dividing by
and taking the square root, which leads to an obvious systematic
scatter increasing fractionally to lower luminosities i.e. to smaller radii.
There are further factors leading to uncertainty in the observed values in
Fig. 10. These are: (a) faint regions are difficult to detect against the
true diffuse background of HH
within the galaxy; this is especially so
within the arms and the bar, though less of a problem in the interarm disc. (b) Overlap may
occur especially in the arms and the bar, which may cause significant underestimation of the
volume.
In Fig. 10, we have plotted HII regions of the bar and of the disc, using different symbols.
There is no clear difference between the properties of the two sets of regions characterized by
this plot.
Copyright The European Southern Observatory (ESO)