Spectacular black smokers along the mid-ocean-ridge crest represent a small fraction of total hydrothermal heat loss from ocean lithosphere. Previous models of measured heat flow suggest that 40%-50% of oceanic hydrothermal heat and fluid flux is from young seafloor (0.1-5 Ma) on mid-ocean-ridge flanks. Despite evidence that ridge-flank hydrothermal flux affects crustal properties, ocean chemistry, and the deep-sea biosphere, few ridge-flank vent sites have been discovered. We describe the first known seafloor expressions of hydrothermal discharge from tectonically formed abyssal hills flanking a fastspreading ridge. Seafloor manifestations of fluid venting from two young East Pacific Rise abyssal hills (0.1 Ma at 10؇20N, 103؇33.2W; 0.5 Ma at 9؇27N, 104؇32.3W) include faultscarp hydrothermal mineralization and macrofauna; fault-scarp flocculations containing hyperthermophilic microbes; and hilltop sediment mounds and craters possibly created by fluid expulsion. These visible features can be exploited for hydrothermal exploration of the vast abyssal hill terrain flanking the mid-ocean ridge and for access to the subseafloor biosphere. Petrologic evidence suggests that abyssal hills undergo repeated episodes of transitory fluid discharge, possibly linked to seismic events, and that fluid exit temperatures can be briefly high enough to transport copper (Ն250 ؇C).Figure 1. A: Locations of East Pacific Rise (EPR) abyssal hill hydrothermal sites. B: Alvin depth profile across abyssal hill at Tevnia site; labeled areas show location of vent fauna and mineral deposits (Fig. 2). C: Alvin depth profile across axis-facing fault scarp of abyssal hill at MM site; labeled areas show where microbial floc (Fig. 3) and sediments (Fig. 4) were collected. for 1994 dive observations, D. Valentine, B. Popp, and E. Olson for help with CH 4 analyses, T. Hill for C isotope analysis, J. Blasius for aid with core sampling, N. Ward and J. Heidelberg of The Institute for Genomic Research for help with microbial sample collection at sea, and P. Holden, M. Lamontagne, and J. Saleta for aid with microbiological analyses at

[1] Heat flow data indicate that most hydrothermal heat loss from ocean lithosphere occurs on the flanks of the mid-ocean ridge, but few ridge flank hydrothermal sites are known. We describe the first nonseamount, abyssal hill hydrothermal mineral deposits to be recovered from the fast spreading East Pacific Rise (EPR) flanks. Deposits were sampled at two sites on an abyssal hill $5 km east of the EPR axis, just north of Clipperton Fracture Zone at 10°20 0 N, on $0.1 Ma lithosphere. ''Tevnia Site'' is on the axis-facing fault scarp of the hill, and ''Ochre Site'' is located $950 m farther east near the base of the outward-facing slope. Clusters of fragile, biodegradable Tevnia worm tubes at both sites indicate that hydrothermal fluids carried sufficient H 2 S to sustain Tevnia worms, and that fluid flow waned too recently to allow time for tube destruction. Presence of microbial mats and other biota also are consistent with recent waning of flow. The deposits are mineralogically zoned, from nontronite-celadonite to hydrous Fe-oxide+opaline silica to Mn-oxide (birnessite and todorokite). This places them into a distinctive class of Fe-Si-Mn hydrothermal deposits found along tectonic cracks and faults in young oceanic crust, and suggests that (1) deposits precipitated along an O 2 gradient between ambient seawater and hydrothermal fluid; (2) fluid temperatures were <150°C; and (3) undiluted fluids were Mg-depleted, and Fe-, K-, Si-and Mn-enriched. These fluids may derive from high temperature seawater-basalt interaction ± phase separation proximal to the axial melt zone, and lose Cu and Zn before venting due to conductive cooling and/or pH increase. Ochre Site samples are purely hydrothermal; however, Tevnia Site samples incorporate volcanic, sedimentary, and fossil components, and exhibit at least three generations of fracturing and hydrothermal cementation. The Tevnia Site breccias accumulated on the exposed fault scarp, possibly during multiple slip events and hydrothermal pulses as the abyssal hill was uplifted. We hypothesize that frequent earthquakes rejuvenate young abyssal hill hydrothermal systems episodically over 10 4 -10 5 years, tapping axial heat and hydrothermal fluids, sustaining biota, and likely helping to chill the margins of the axial melt zone.