A water fountain and three associated pools are perched on the wall of the Volcan Alcedo caldera, at the end of a line of active fumaroles. In July 1970, the lower basin was dry and the overflow pool was partially filled; the fountain pool was nearly full and was at the boiling temperature. Gases forcefully pumped from a vent caused a continuous fountaining and abundant steam. Opal shoreline deposits and “mushrooms” mark former maximum water levels, during which the overflow pool level is controlled by a spillway to the lower basin. A levee between the fountain and overflow pools causes their separation when the water level drops. Reports show that the pools were full in August 1968, became completely dry by October 1970, and partially refilled in 1971, Throughout the cycle, sulfurous steam flowed from the vent. The water chemistry of the fountain and overflow pools shows an increase in dissolved constituents in the latter; both are high in boron. Calculations indicate that the water of the overflow pool is not in equilibrium with its deposits, but is equilibrated with the atmosphere. The fountain pool shows the opposite conditions. Silica content and the location of geyserite deposits indicate that opal saturation occurs as the water cools and circulates through the overflow pool. Strata in the levee show cycles of mud, opal, and sand deposition which correspond to periods of maximum filling, partial filling, and dryness, respectively. Subaerial deposition of the sand is shown by its high physical maturity (rounding and frosting) and lack of chemical alteration. Sand fountaining and the formation of a sand cone, associated with a high-pressure fumarole, were observed at another area in the Galapagos. The physical nature of this sand is nearly identical to that of the levee around the Alcedo vent. The same process is proposed for the formation of the levee deposits; this sub-aerially formed cone is modified during periods of pool filling. A turbulent gas flow velocity of about 6 m per sec would account for the position and size of the levee sand grains. A very small driving force at depth would be required. Gas analyses from fumaroles north of the fountain have a relatively high magmatic component and favor the stability of magnetite. A similar gas, with periods of slight oxidation, would be compatible with the chemistry of the levee sand. The source of water in the pools is largely meteoric; the fountain and overflow pools depend more upon subsurface flow, while the lower basin receives surface water from the caldera wall.