Background

Ionized gas forms an important component of the interstellar medium in late-type galaxies. The bulk of the emission comes from H II regions, but a significant fraction comes from regions far removed from these ionising clusters. This emission is distributed in loops and filamentary structures, and a component that appears diffuse.

Spectroscopy of the filamentary and diffuse components in other galaxies (Hunter et al 1993, Ferguson et al 1996, Hunter & Gallagher 1997, Martin & Kennicutt 1997) show this emission is predominantly photoionised, and the most likely source of ionisation is the UV continuum from bright starforming regions, even though the projected path lengths are of order 1 kpc. This implies the giant H II regions are not ionisation-bound, and that the interstellar medium beyond the H II region environs must be rather porous. How porous the ISM must be depends on the (unknown) location of the ionized emission relative to the disk.

The diffuse emission is likely akin to the Reynolds layer in our own galaxy, where many small holes are sufficient for radiation escaping the midplane to keep the gas at large scale heights ionized (Miller & Cox 1993, Dove & Shull 1994).

The loops and filaments can be divided into two classes, one where the structures at least partially encircle young clusters, and another where the filament size scale is often large ( kpc) and there is no clear association with an ionizing cluster (Hunter & Gallagher (1997) call these `superfilaments'). The former are probably within the disk plane and can be understood as classical expanding shells with the interior largely cleared of H I by the expansion and heating processes, and the filamentary ionized emission coming from the interior of the shell wall. For the latter class, the emission might be part of an expanding-shell shock front, or an ionization front on a relatively stationary H I structure. Assuming these superfilaments to be photoionised, if they lie within the host's disk the column density along the path between the emission region and ionising cluster must be rather low. The H I deficit should be visible in radio synthesis observations. If the emission comes from gas above the disk plane, the radiation transport problem is similar to that of the diffuse gas, and the the disk may appear more nearly uniform.

In the accompanying images I present comparisons of ionized and neutral gas morphology with the aim of understanding the origin of the large-scale filaments in NGC 4214, and discuss the neutral hydrogen distributionaround the most UV-luminous cluster in the galaxy, NGC 4214#1.

© Copyright Astronomical Society of Australia 1997