The Silsilah tin deposit (lat 25 degrees 06' N, long 42 degrees 40' E) consists of a group of pervasively greisenized, flat-topped granite cupolas within a 12-km-diam ring complex. The greisens contain varying amounts of disseminated cassiterite and wolframite. Several types of quartz veins are peripheral to the greisens; some of these contain minor wolframite. The deposit is genetically associated with a highly differentiated, peraluminous alkali-feldspar granite (587 + or - 8 Ma) that is part of a mostly peralkaline, igneous ring complex intruded into Late Proterozoic, immature sandstones of the Murdama Group. We recognize four distinct phases of the peraluminous granite. Only the smallest, most highly differentiated cupolas contain significant tin greisen mineralization. Greisens developed beneath aplitic carapaces that overlie the granite and created impermeable barriers to rising volatiles. The geometry of a cupola correlates strongly with the intensity of alteration; cupolas with the smallest cross sectional areas and steepest marginal contacts have the most intensely greisenized apexes. The paragenetic sequence can be divided into five stages: pegmatite formation, locally pervasive albitization, locally pervasive greisenization and deposition of cassiterite, deposition of quartz-wolframite veins, and deposition of quartz veins with minor base metal sulfides. Pressure-corrected fluid inclusion filling temperatures indicate that the hydrothermal system generally cooled as it evolved and that the delta ¹⁸O values of the hydrothermal quartz increased from 10.8 to 15.7 per mil. Calculated delta ¹⁸O values of the hydrothermal fluid varied concomitantly from the pegmatite stage (delta ¹⁸O fluid approximately 8.6ppm; T [asymp] 550 degrees C) to the greisen stage (5.4 and 5.6[ppm; T [asymp] 360 degrees C), the quartz-wolframite vein stage (6.3 and 7.5ppm; T [asymp] 390 degrees C), and the late vein stage (4.0 and 5.1ppm; T [asymp] 270 degrees C). This evolution probably reflects the admixture of generally increasing amounts of meteoric or formation water having a lower delta 18 O value into the cooling magmatic hydrothermal system. In delta ¹⁸O-delta ¹⁸O plots for mineral separates from fresh to altered samples of the peraluminous granite the values for quartz and feldspar conform to a steep, positive-sloped disequilibrium trend that indicates interaction with high ¹⁸O hydrothermal fluid, mirrored by a negative-sloped disequilibrium trend for quartz and mica. These complementary trends suggest strongly that individual granite cupolas were essentially closed systems during alteration. To our knowledge, this is the first oxygen isotope demonstration of a closed-system, hydrothermal metal deposit. The sandstone country rock has whole-rock delta ¹⁸O values of 12.4 + or - 2.0 per mil. The highest values (>13ppm) form a approximately 3-km-wide high delta ¹⁸O annulus immediately peripheral to the ring complex. The data suggest that the country rocks were pervasively exchanged with an outward-migrating, high delta ¹⁸O fluid moving down a temperature gradient. This fluid was probably formation water that flowed radially inward toward the pluton at a deeper level, forming a largely horizontal, unicellular flow system that had fluid-flow lines nearly vertical next to the intrusion, and directed upward and outward at higher levels. Low delta ¹⁸O fluids that formed quartz associated with the tin-tungsten mineralization and later veins are inferred to be the result of a subsequent and distinct phase of the hydrothermal system. Almost all the geologic processes necessary for the formation of an economically viable tin deposit occurred at Silsilah, but the lack of a strong localizing mechanism for cassiterite mineralization resulted in an economically marginal deposit. If the greisenized cupolas had been vertically stacked, as are the intrusions and ore zones in Climax-type molybdenum deposits, or if the mineralizing fluids had been channeled into veins, as in the tin deposits in Cornwall, England, a higher grade deposit might have formed. The generally closed-system behavior of the hydrothermal system at Silsilah may have prevented additional scavenging of metals and the formation of a larger, richer deposit.