5.3. The evolution of the IC gas
Many years before its detection, Limber  had argued that IC gas must exist because galaxy formation cannot be 100 % efficient, and that it must evolve through the loss of gas from colliding galaxies. The IC gas was eventually detected [299, 201] in 1971. In those years, Gott & Gunn [188, 200] developed their theory of intergalactic gas infall into clusters. They argued that this infall could generate a hot IC gas through shock heating. They suggested that irregular clusters are seen in a pre-collapse phase, so that their IC gas had not yet reached high temperatures. In this way they hinted at the existence of a class of X-ray faint clusters (which are now being discovered, see Holden et al. ). Gunn & Gott  also suggested ram pressure as a mean to strip gas from cluster galaxies and enrich the IC medium.
An early gas infall became a common feature of models in which the IC gas is in hydrostatic equilibrium in the cluster gravitational field (Lea , Gull & Northover , Cavaliere & Fusco-Femiano [98, 99]). On the other hand, Yahil & Ostriker  developed a theory with an IC gas outflow. They argued that the gas shed from the galaxies would feed an outflow wind from the cluster. Such a radial outflow of the IC gas was soon found to be at odds with the random direction of the cluster galaxy radio-tails (Lea ).
In 1973 Lea et al.  remarked that since the mass of IC gas is comparable to the total mass in cluster galaxies, not all of the IC gas can originate from cluster galaxies, and most of it must be primordial. On the other hand, Larson & Dinerstein  advocated for a galaxy origin of a significant fraction of the IC gas, through supernova explosions and stellar winds. Their model predicted a significant abundance of heavy elements in the IC gas. The hydrodynamic numerical simulations by Lea & De Young  indicated that as much as 90 % of the gas can be removed from galaxies moving through the IC gas at transonic speed.
In 1977, Iron was found in the IC gas [309, 308, 410], proving that at least some of the IC gas had been processed in stars. A purely primordial origin of the IC gas was thus ruled out. As a matter of fact, observations seemed to indicate that the IC Iron mass was larger than could be produced in cluster galaxies. This led Vigroux  to suggest an early heavy-element enrichment of the IC gas by a pre-galactic population of massive stars. Fabian & Pringle [158 noted however that the estimates of the total cluster Iron mass were very uncertain, being based on extrapolations from the inner regions. Recently, Gibson & Matteucci  have shown that even a large population of dwarf cluster galaxies, as implied by the steep cluster LF, could account for the bulk of the IC gas and metals.
Norman & Silk  and Himmes & Biermann  developed models for the temporal evolution of the IC gas. An initial amount of IC gas would first originate from galaxies through supernovæ emission. Only then, ram pressure stripping could start. This model was proposed as an explanation of the Butcher-Oemler effect (see Section 5.2).
In 1980, White & Silk  noted, in disagreement with Gingold & Perrenod , that mergers of subclusters can lead to strong heating of the IC gas in the compression region. This was later observed .
Cowie & Perrenod 's models indicated a mild evolution of the X-ray cluster luminosity with redshift. Perrenod 's more refined model, now including a cluster gravitational potential varying in time, predicted instead a very strong evolution of the X-ray cluster luminosity, a factor ten from z ~ 1 to the present. Perrenod  later showed that the evolution rate of the cluster X-ray luminosities was related to the density of the Universe, so that X-ray observations of distant clusters could be used to put useful cosmological constraints.
Perrenod's prediction of a strong evolution in the cluster X-ray properties was first tested observationally by Henry et al. . Unfortunately, the wide range of X-ray luminosities for distant clusters made it impossible to test the model. Two years later, in 1981, Perrenod & Henry  argued for an X-ray temperature negative evolution with redshift, based on a limited sample of seven clusters observed at z > 0.1. Such an evolution was however not confirmed in other investigations. First, White et al.  detected an extremely bright and hot X-ray cluster at z = 0.54, then, Henry et al.  did not detect any change in the slope of the cluster X-ray luminosity function with redshift.
The first observational evidence for a cosmological evolution of the X-ray cluster properties dates back to 1982. Anticipating the results that were to be published in their entirety by Gioia et al.  many years later, Stocke et al.  noted that the clusters detected in the flux-limited Einstein Medium Survey Sample have a low average X-ray luminosity and a low average redshift, and their total number is half that expected for a uniform distribution of sources. This was interpreted as evidence for a negative evolution of the cluster X-ray luminosity function.
This evolution is now confirmed for the high-luminosity tail of the X-ray clusters only (see MULLIS, these proceedings). The high fraction of hot X-ray clusters at high redshift is now considered to be a strong evidence for a low-0 Universe (see GIOIA, these proceedings).