The much greater spread in the distribution of HI surface densities among the early type galaxies remains a puzzle. While some E and S0 galaxies are undetectable in the HI line at a sensitivity level corresponding to as little as 106 M, others contain a relatively large amount of neutral gas (Wardle and Knapp 1986). The case of the S0 galaxies is particularly relevant since we have seen that the HI deficient galaxies could evolve into S0-like objects in X-ray clusters. In the context of identifying the high redshift blue cluster galaxies as the stripped progenitors of today's cluster S0's, Larson et al. (1980) have proposed that S0's in the cluster environment were stripped of their outer gas reservoirs at an early epoch so that star formation and continued disk growth were arrested early on. This process would continue to strip any disk object that passed through the core of a rich cluster. Under this scenario, S0 galaxies in low density regions might still possess some gas, and thus a segregation into HI-normal and HI-poor objects would be expected. In other cases, most notably those of the polar ring or spindle S0's, authors have invoked an external origin for the gas, probably through the accretion of a low-mass gas-rich neighbor (van Driel 1987). The recent acquisition of gas is particularly appealing to explain not only the polar rings but also the annular distribution of HI seen exterior to the optical disk in some S0's.
In an early study of the HI in S0 galaxies in the Virgo cluster, Giovanardi et al. (1983b) applied the "IN versus OUT" technique to show that early type spirals - S0, S0/a and Sa's - outside the Virgo core were more likely to be detected in HI than were similar objects within the core. This finding was more quantitatively confirmed by Chamaraux et al. (1986). Thus, it seems that the S0 galaxies follow the trend noted by Dressler (1986) that early-type objects are even more HI poor than later spirals.
In an attempt to examine the cause of the greater gas depletion among early type spirals, Haynes et al. (1989) have examined the HI and far infrared emission in samples of relatively bright S0, Sa and Sc galaxies in the Local Supercluster. For the early type objects, new high sensitivity HI observations were conducted of previously undetected or unobserved objects using either the Arecibo 305m or the late N.R.A.O. (2) 91m telescope whichever was appropriate. Coaddition of the Infrared Astronomical Satellite IRAS far infrared observations were made using the Infrared Processing and Data Analysis Center (3). This effort has concentrated on an examination of the occurrence of HI deficiency in terms of local environment. In order to test for possible differences in the HI content that are dependent on local parameters, Haynes et al. have calculated two measures of local environment using a complete redshift catalog of the Local Supercluster. The first parameter is a local "density" counted as the number of companions within a sphere of 1 Mpc centered on each sample object. The second attempts to measure the sum of tidal forces caused by objects within the same 1 Mpc sphere and incorporates both the projected separation and relative radial velocity. Within the Local Supercluster, the environments as characterized by these two parameters, vary by a factor of 200 according to this density measure. Segregation is only seen in the highest density bins (the center of the Virgo cluster) where few Sc's reside.
The environmental influences have been tested by the comparison of the distribution functions of HI content for populations representing the high density and low density regions. Non-parametric statistical techniques that make use of non-detections have been employed, since despite the increased sensitivity of the new observations, many S0 galaxies still remain undetected. Haynes et al. quantitatively confirm the results of previous authors that the S0 galaxies in the highest density regions of the Local Supercluster, namely the Virgo core, are characterized by an average HI surface density a factor of four lower than those in the lowest density regions. At the same time, they find no significant difference in the far infrared properties of galaxies in high and low density regions. By extending the analysis to similar samples of Sa and Sc galaxies which represent subsets of the survey of Magri and Haynes (in preparation), they conclude that the depletion of HI in galaxies in high density regions is more severe among both S0's and Sa's than among Sc galaxies. A similar analysis was conducted using the tidal disruption parameter but, unlike in the density case, no statistically significant difference in the HI content was seen in high damage cases versus low damage ones. The distribution of the tidal parameter for a given value of the density parameter is quite large and we believe this result occurs because of the uncertainty of projected separations, radial velocities, orbits and angular momentum vectors and the higher velocity dispersion in clusters, for which the disruption parameter as defined will be reduced for collisions of the same projected impact parameter. Interestingly, the most gas-rich S0's can all be explained in terms of recent accretion events.
It should be kept in mind that there is a wide range in the HI and FIR surface densities seen among the S0's. While some seem to possess interstellar properties typically associated with mid spirals (the "closet" Sb's), others are extremely gas and dust poor. While the mechanism for gas removal is yet unclear, it seems likely that the processes that affect spiral galaxies passing through the cores of rich clusters will also affect the S0's that reside there. With a reduction of the gas content and subsequent disk fading, spiral structure must fade. While it is difficult to take a highly flattened Sc disk and turn it into a bulge-dominated S0 (although see Farouki and Shapiro 1980), it may be possible to turn early spirals (Sa's) into S0's.
2 The National Radio Astronomy Observatory is operated by Associated Universities Inc. under contract from the National Science Foundation. Back.
3 The Infrared Processing and Analysis Center is operated by the California Institute of Technology for the National Aeronautics and Space Administration. Back.