Current evolutionary theory indicates that evolving stars in globular clusters arrive on the horizontal branch with ~30% less mass than they had on the main sequence. If, as seems likely, this mass loss results from the outflow of unprocessed material at the stellar surface during the giant stage, and if the ejected mass were retained within the cluster between successive sweeps through the galactic plane (~108 yr), sufficient hydrogen (100-2000 M⊙ should accumulate for detection. Radio searches, however, have failed to find evidence for either neutral or ionized gas. The hydrogen mass upper limits obtained by various workers are: Hills and Klein (1973) ionized H, 3.8 cm, <20-200 M⊙; Knapp, Rose and Kerr (1973) neutral H, 21 cm, <1-20 M⊙; Conklin and Kimble (1974) neutral H, 21 cm, <0.7-1.4 M⊙.
We have extended the search into the optical region and have also calculated time-independent gas flow models to try to resolve this anomaly.