cooler, diffuse hydrothermal fluids directly from the basaltic seafloor, rather than from hydrothermal mineral constructions; (9) rapid and extensive growth of flocculent white bacterial mats (species unknown) on and under the seafloor in areas experiencing widespread venting of diffuse hydrothermal fluid; and (10) subseafloor downslope migration of magma normal to the ridge axis in a network of small-scale (1-5 m diameter) lava tubes and channels to distances at least 100-200 m outside the ASC. We suggest that, in April, 1991, intrusion of dikes in the eruption area to < 200 m beneath the ASC floor resulted in phase separation of fluids near the tops of the dikes and a large flux of vapor-rich hydrothermal fluids through the overlying rubbly, cavernous lavas. Low salinities and gas-rich compositions of hydrothermal fluids sampled in the eruption area are appropriate for a vapor phase in a seawater system undergoing subcritical liquid-vapor phase separation (boiling) and phase segregation. Hydrothermal fluids streamed directly from fissures and pits that may have been loci of lava drainback and/or hydrovolcanic explosions. These fissures and pits were lined with white mats of a unique fast-growing bacteria that was the only life associated with the brand-new vents. The prolific bacteria, which covered thousands of square meters on the ridge crest and were also abundant in subseafloor voids, may thrive on high levels of gases in the vapor-rich hydrothermal fluids initially escaping the hydrothermal system. White bacterial particulates swept from the seafloor by hydrothermal vents swirled in an unprecedented biogenic 'blizzard' up to 50 m above the bottom. The bacterial proliferation of April, 1991 is likely to be a transient bloom that will be checked quickly either by decline of dissolved gas concentrations in the fluids as rapid heat loss brings about cessation of boiling, and/or by grazing as other organisms are re-established in the biologically devastated area.