ISM MODELS. We organize the various ism models into catagories. We include some global that are very comprehensive and some that are more limited and make some comments. 1. Models in which the state of matter is based primarily on thermal processes. ***Field, Goldsmith, Habing (FGH, 1969 ApJ 155 L149). The original global model of the ISM considered two phases, the CNM and the WMN, in hydrostatic equilibrium in the galactic z-gravitational field. Microscopic thermal equilibrium, heating=cooling; no nonequilibrum states. Added features included a magnetic field pressure component to product the observed z-thicknesses and CNM cloud motions to produce the observed cloud-cloud velocity dispersion. ***McKee/Ostriker (MO, 1977 ApJ 218, 148). SN carve out volumes in space forming the HIM, which is not in thermal or ionization equilibrium. The rest of the space is occupied by clouds with CNM at in middle, WNM on periphery, WIM on outside. WNM heated by soft XR from HIM and thus has relatively high ionization; can produce significant pulsar dispersion and Halpha emission. Except for HIM, microscopic thermal equilibrium, heating=cooling; no nonequilibrum states. HIM evaporates these clouds, convertng warm/cool gas to HIM. The SN shock sweeps up ambient HIM; shells become radiative and produce warm/cold media. Note statistical equil between these processes. HIM has large scale height and fills the halo. Ambient pressure of ISM is that inside a SNR when it overlaps other SNR. Predicted to be about 4000 cm^-3 K. Applies to all phases, which are in pressure equilibrium. Pressure fluctuates from place to place, depending on SN proximity in space/time. Most SN energy should be radiated at soft XR. General picture: ISM is swiss cheese, with the cheese being the HIM and having dominant filling factor. ***Superbubble variant: see McKee review. Halo filled with hot gas by big superbubbles breaking through the disk. Galactic fountain: hot gas up, cold 'rain' down (Shapiro/Field (1976 ApJ 205, 762). ***Cox/Slavin (1992 ApJ, 392, 131). Include magnetic fields and enhanced cooling to argue that HIM filling factor is low, and that superbubbles don;t every break out of the disk to fill the halo with hot gas. For arguments see Cox review. ***Time dependent models e.g. Gerola, Kafatos, McCray 1974 ApJ 189, 55. In original form, SN produce XR burst, which impulsively ionizes and heats surrounding gas out to some large radius. Gas cools, recombines, then is randomly hit by another SN shock. Thermal state of gas is probabilitic with no unique relation among temp, ionization, density. Current idea: replace SN by soft Gamma-ray repeaters as XR source (totally undeveloped). ---------- 2. Models in which the state of matter is based on dynamical processes (as primary agents) and thermal as secondary agent. ***Wada/Norman (2001 ApJ 547, 172). Gravititational instabilities with spiral density waves (on the largest scales) and SN on smaller scales (including star formation feedback) produce a wide distribution of density, temp; the distributionis continuous and the term 'phase' is not useful. HIM ranges from 1e6 to 1e8 K and occupies most of the volume. Some large holes are a result of nonlinear dynamics, not explosions. Stellar formation is positive feedback, resulting in large fluctuations in star formation rate (starbursts explained). ***Vazquez-Semadeni, Gazol, Scalo (2000 ApJ 540, 271) perform numerical simulations of turbulence including magnetic fields and stellar feedback but neglecting large scale Galactic dynamimcs. Empohsize smaller scale structure. No microscopic thermal equilibrium, particularly for WNM. ***Walters/Cox (2000 ApJ 549 353). Large-scale explosive events near z=0 (spiral density wave shocks) produce oscillatory modes in z. Gas goes up, falls back; gas at high z falls later and faster, colliding with slower moving gas producing shocks with characteristic velocity 50 km/s, explaining the 'intermediate velocity gas' so prevalent at high b near the sun. at high z, shocks go faster and produce the hot halo gas. smaller scale perturbations, e.g. SN, produce smaller waves and produce multiple narrow velocity components in interstellar velocity spectra. most lines are produced by gas that is almost cospatial. Probable extensions to other aspects of the ISM are possible. Impressive is that this explains a major features in neutral gas ISM spectra and, also, z-distribution of the 'HIM'. ------------ 3. random comments. MO predict the interstellar pressure from the SN rate, without invoking Galactic hydrostatic equilibrium. Yet the interstellar pressure is determined by hydrostatic equilibrium! So should we conclude that the SN rate is tied, through the pressure, to Galactic hydrostatic equilibrium? And if so, how does Toomre's famous Q parameter tie in? When models are examined from this more global perspective, it leads to an expectation of feedback produces tying all processes together. Complications are frightening! Wada/Norman is, in my view, a first attempt to synthesize all this--and they fiund that the feedback is positive, producing unstable equilibrium. Galactic SN rate decreases considerably with increasing radius. Also, thickness of gas layer increases, and mean density at z=0 decreases, considerably with radius. Net result is that FGH may apply in outer Galaxy, MO in inner--with appropriate modifications if Cox/Slavin are right... OVI: MO predict about 10 X more interfaces (number of independent velocity components) a than Jenkins found; Cox/Slavin predict fewer. But then Shelton/Cox (1994 ApJ 434 599) reinterpreted Jenkins to find fewer OVI components! (note the power of theorists doing observational interpretation!). A basic difference between Slavin/Cox and MO is in whether the SN energy goes into heating gas (and is thus lost to radiation) or compressing magnetic field (which springs back). MO predict more HIM emission than SC--not seen in external galaxies. But most of MO's HIM is at 5e5 K, too cool to emit XR and to warm for OVI absorption: hard to detect! Superbubbles versus supernovae: if enough SN are clustered so that individual SN are unimportant, MO needs modification. McKee claims this is not the case. What are the roles of Parker instability and cosmic ray pressure? How do the magnetic field and cosmic rays enter in ISM models? As far as I know, the only discussion of hydrostatic equilibrium including these are Badhwar and Stephens (1977 ApJ 212, 494) and Boulares and Cox (1990 ApJ, 365, 544). Parker (1990 IAU SYMP 140, 169) estimates energy input to the Balactic halo from 'Galactic flares' (reconnection of magnetic fields in Parker-instability coronal-like loops in the halo) of 1e40 to 1e41 erg/s, which is stated to be comparable to XR power observed from halos of external galaxies. Energy input would be localized, as in solar flares. It is equivalent to 1 sn per 1000 yr, about 20 times smaller than the actual rate--but what fraction of SN energy puts gas into the halo?.