Isotope geochemistry of oxygen in the sedimentary sulfate cycle

Isotope geochemistry of oxygen in the sedimentary sulfate cycle

Chemical Geology, 25 (1979) 1--17 1 © Elsevier Scientific Publishing Company, Amsterdam -- Printed in The Netherlands I S O T O P E G E O C H E M I...

1MB Sizes 2 Downloads 17 Views

Chemical Geology, 25 (1979) 1--17


© Elsevier Scientific Publishing Company, Amsterdam -- Printed in The Netherlands



T. H O L S E R

I , I S A A C R. K A P L A N

2 , H I T O S H I S A K A I 3 and I S R A E L Z A K 4

iDepartment of Geology, University of OregOn, Eugene, O R 97403 (U.S.A.) 2 Department of Earth and Space Sciences and Institute of Geophysics and Planetary Physics, University of California, Los Angeles, C A 90024 (U.S.A.) 3Institute for Thermal Spring Research, Okayama University, Misasa, Tottori-ken 682-02 (Japan) 4 Department of Geology, Hebrew University, Jerusalem (Israel) (Received April 7, 1978; revised and accepted September 5, 1978)

ABSTRACT Holser, W.T., Kaplan, I.R., Sakai, H. and Zak, I., 1979. Isotope geochemistry of oxygen in the sedimentary sulfate cycle. Chem. Geol., 25: 1--17.

A re-evaluation of the isotopic geochemistry of oxygen involved in the exogenic sulfur cycle leads to a model in which 6 lSOso 4 of seawater is determined by a dynamic balance of four fluxes: inputs by erosion of evaporite sulfate and by oxidative erosion of sulfide, and outputs of sulfate into evaporites and of carbon dioxide into seawater by bacterial reduction of sulfate to sulfide. The sulfate deposition--erosion loop is closed, but the sulfide loop is open, connected through the fixed reservoir of laOH o of seawater. The level of 6 ISOso is apparently not appreciably affected by either equ~ibration with laOH~o, or by Lloyd#s proposed fast reduction--oxidation cycle on the sea floor. The possible effect of equilibration during a subsea hydrothermal circulation is unclear from available data. Calculations of hypothetical equilibration of l aO/'~Oso4 with 180/1~OI~o, either from the dehydration of gypsum or from formation waters, give ~ laOso4 values much higher than any observed in old evaporites. Consequently, these processes were probably not significant in altering the 61,Oso4 of seawater that is recorded in evaporites.

INTRODUCTION Extensive n e w d a t a o n ~ ~SOso ' in sulfate minerals o f various geological ages - - t h e " o x y g e n i s o t o p e age curve " (Sakai, 1 9 7 2 ; C l a y p o o l et al., 1972, and in prep.) - - can o n l y be i n t e r p r e t e d after a critical evaluation o f the isotopic g e o c h e m i s t r y o f o x y g e n w h i c h influences the sulfur cycle. This p a p e r will a t t e m p t t o evaluate this c y c l e based on available evidence. T h e overall isotopic g e o c h e m i s t r y o f o x y g e n has b e e n s u m m a r i z e d b y Garlick ( 1 9 6 9 ) and b y Hoers *Publication No. 1794: Institute of Geophysics and Planetary Physics, University of California, Los Angeles, CA 90024, U.S.A.

{1973), but those authors give only cursory attention to the interaction of oxygen with sulfur. ISOTOPIC GEOCHEMISTRY OF OXYGEN IN THE SEDIMENTARY CYCLE

Many of the features of the present sulfur--oxygen cycle that contribute to the variation of its isotopes are diagrammed in Fig. 1. Seawater is by far the largest oxygen reservoir, at 81SOlo (SMOW) = 0°/00• Variations of 5 'sO in the sea due to ice removal during glacial periods are less than +1.0°/00 (Emiliani and Shackleton, 1974). A long-term change has been proposed on the basis of analyses of SiO2 oxygen in marine cherts {Chase and Perry, 1972; Perry and Tan, 1972), but more recent and detailed work (Knauth and Epstein, 1976) has ascribed the variations in chert to climatic changes (and temperature has an insignificant effect on the 5 'SOso ' to be considered, e.g.: Table I, column 3). Hydrothermal buffering at ocean ridges has been postulated by Muehlenbachs and Clayton (1976) to keep seawater near 8 'sOH2 o = 0%0. Consequently the ~ 'sOH2o in seawater may be assumed constant for the Proterozoic--Phanerozoic time with which we are concerned. Reservoirs tied to the seawater value are similarly controlled in time: carbon dioxide in the atmosphere and bicarbonate in seawater are held at 81sO = +41 and +30°/00, respectively, by exchange equilibria; fresh water is variably negative with respect to seawater due to evaporation kinetics (Garlick, 1969). In the continental biological cycle, transpiration in leaves and respiration (mainly by soil bacteria) combine to boost 8 'sO of the resulting O2 gas up to about +32% o, but in the sea the lack of transpiration and the lower temper50



20 6180



Fig. 1. Oxygen isotope geochemistry, showing the relations among 8 1sO levels in various reservoirs that have some concern with the sulfur cycle. Double-ended arrows are meant to indicate substantially equilibrium fractionations, single-ended arrows indicate other (kinetic) relations. The dashed-line relations indicate hypothetically calculated isotope alterations, which are in fact not observed (see text). Sources and reliability (highly variable) of the information are discussed in the text.

" ~ o

a~ m o

~o ~


l'O O'J 0,] ~'~ + ,.1., ~o







b~ M

a "~



~0 cO o'J 0,] ~-4







,.14 o-~ L-,. 0

,~ '0


00~ ~,] ,...

I M ~ _ ~ S . O ~ O



Fig. 2. A simplified main cycle of sulfur in the sedimentary cycle, and for oxygen involved in the sulfur cycle (after Claypool et al., in prep. ). Reservoirs of mass M and isotope level a are interrelated by fluxes dM with isotope differences (fractionations)z~. ~ is the fraction of river sulfate derived from the surface oxidation of sedimentary sulfide.

atures of respiration hold this to 6 '80 = +10%0 ; mixing of these two sources gives an average for the present atmosphere of a ' 8 0 = +23%0 (Wagener, 1973). A second key point of Fig. 1 refers to sulfate in seawater with a present isotope composition of 6 ~8Oso' = +8.6°/oo, that does n o t represent equilibrium with 618Oa2 o of seawater (Lloyd, 1967, 1968). In the following discussion we account for this value as a result of dynamic equilibrium of input and output processes. One consequence of this dynamic balance is its possible variation with time, in explanation of the oxygen-isotope age curve. Figure 2 displays the inputs and outputs of sulfate in seawater as fluxes dM and fractionations A of both O and S between reservoirs M with isotope levels 6. Refer to this figure for symbolism used in the following discussion. Consider first the inputs to the sea. Oxidative weathering of sulfide ore deposits resulted in 640 = --1 to - - 6 0 o (Sakai and Matsubaya, 1974; Smejkal, 1978). Although the air and water from which the sulfate was precipitated were n o t analyzed in this study, the 618Oso ' observed therefore suggests a participation of both water oxygen and air oxygen in the process of oxidation of the sulfide at the surface. This may be compared with the experiments of Lloyd (1967, 1968), who oxidized a sodium-sulfide solution with oxygen and interpreted his results in terms of the reaction: S 2- + H20 + 02 -~ (SO3) 2-


with the participation of one water oxygen and two gaseous oxygen atoms. The molecular oxygen u ~ d in thereaction was found in Lloyd's experiment to have been fraetionated by 4 ! sO = --8.7%0 from the original oxygen. Further oxidation of sulfite to sulfate was found to use only gaseous oxygen, but Lloyd does not state whether this molecular oxygen was also fractionated. Lloyd also observed that oxidation of sulfite was in competition ~with a fast equilibrium of sulfite oxygen w i t h water oxygen, but no data seem to be available on any fractionation invol~)ed in that equilibration. Furthermore, the dominant sulfide that must be oxidized in the sulfur cycle is pyrite; Nakai (1963) has shown that the oxidatio~ of pyrite in nature is d o m i n a n t l y by sulfur-oxidizing baeteria, and the produet of.:Sueh baeterial oxidation is sulfate, although the intermediate steps are complex and as yet not identified (Goldhaber and Kaplan, 1974). Mizutani and Rafter (1969) observed that in bacterial oxidation

of native sulfur, the sulfate formed shows the same ~ 1sO value as the water in which it was oxidized. However, no laboratory isotope measurements on oxidation of pyrite have been reported. In view of these complications, we assume in Figs. 1 and 2 a nominal value of A210 = -2°/oo derived b y some combination of fractionated air oxygen and fresh-water oxygen, with the understanding that this is probably variable and certainly imprecise. Pre-Devonian air oxygen should lack the contribution of land plants and hence should be lighter (81sO = +10%0) than air oxygen t o d a y (Wagener, 1973); however, no such deviation is observable in our isotope age curve of oxygen in sulfate (Claypool et al., in press) at the time land plants became important. Mixing of the sulfide-generated sulfate with an approximately equal a m o u n t (Berner, 1971; Garrels et al., 1973) of dissolved evaporite sulfate with a mean value of, say 8sO = +12°/00 (Sakai, 1972; Claypool et al., 1972, and in prep.), might yield a mean for sulfate carried b y rivers into the sea of perhaps 540 = +5°/oo. Measurements of 8180 in sulfate of modern'rivers are few (Longinelli and Cortecci, 1970; Schwarcz and Cortecci, 1974), and even in isolated localities these u n d o u b t e d l y suffer from a large b u t indeterminate fraction of contamination b y industrial atmospheric sulfur (Kellog et al., 1972). In the absence of firm data, this value of ~ 40 = + 5 ° o for pre-industrial rivers must remain a guess. Now consider the o u t p u t s from the sea. No definitive measurements of the fractionation of oxygen during sulfate reduction have been made in natural systems. In laboratory experiments of various kinds it is well known, in the case of sulfur, that the fractionations measured are substantially below those that can be deduced for natural systems, probably because of the great difference of rates involved (Goldhaber and Kaplan, 1974). Consequently in estimating the fractionation of oxygen during the reduction of sulfate in natural systems, we take the ratio of fractionations of oxygen and sulfur measured in laboratory experiments involving the reduction of sulfate, and then multiply this b y the fractionation of sulfur observed in natural systems. First we note that the fractionation of oxygen during sulfate reduction,/~ 120, has been measured in laboratory experiments that approximate the open system at the top of the sediment, at a ratio of ~ 12S/~ 120 = 4 (Mizutani and Rafter, 1969). Although the fractionation of O is dependent on 518OH20 of the seawater present during the reduction (Mizutani, 1973), the variation is not important within the small range of marine 618OH2o discussed above. The ratio determined by Mizutani and Rafter (1969) has not been checked b y measurements in seawater from cores of o p t i m u m areas for sulfate reduction, although Zak et al. (in prep.), did measure a ratio of a b o u t 2.5 in one deep basin in the Gulf of California where a small amount of'sulfate reduction had occurred. For the present, we accept a ratio of 4, with the recognition that it m a y turn o u t to be somewhat lower. A wide range of values for the fractionation, A 12S, has been found by laboratory experiment, b y analyses of sub-sea cores, and b y the comparison

of sulfate and sulfide ore deposits of the same geological age. Analyses of sub-sea cores are probably the most reliable. Most of the reduction of sulfate to sulfide takes place in the upper few centimeters of sediment in organic-rich, rapidly sedimenting basins (Kaplan et al., 1963; Hartmann and Nielsen, 1969; Goldhaber and Kaplan, 1974; Sweeney and Kaplan, 1978a). Such situations show a fractionation close to ~ 12S = -400/00 (Sweeney and KaDlan, 1978b). The same value was found by Jurio {1976) in an analysis of sediments of the Variscan geosyncline. Rees {1973) has explained how such a large fractionation can be generated. A wide variety of other milieux such as subsurface sediments in these same basins, or organic-poor areas {e.g., deep-sea sediments) display an exceedingly wide range of fractionations (Migdisov et al., 1974; Grinenko et al., 1973; Schwarcz and Burnie, 1973; Zak et al., in prep.). We therefore assume A ~O = --10°/0o. Even as an average for natural fluxes, this estimate may be considerably in error. Figure 1 shows diagrammatically oxygen from reduction of seawater sulfate at 5180 = 8.6 - 10 = -1.4°/00, which in the product form CO2 or HCO3 would, of course, quickly equilibrate with the large reservoir of seawater. The kinetic removal of this light oxygen tends to make the residual seawater sulfate oxygen correspondingly heavy. Selection of a best value for the fractionation on precipitation of gypsum, ,30, also presents some problems. If, as in our Table I, one takes the difference of Lloyd's (1968) experimentally measured fractionation factors for anhydritewater and sulfate (solution)-water, the result is ~ ~30 = 9.5%o at 25°C. But in the same paper Lloyd {1968, p. 6104) says: " T h e data on the equilibrium fractionation factors suggest t h a t at 25°C fractiona~ion between dissolved sulfate and anhydrite might be as much as 6°/oo''. The discrepancy between his equations and his statement is unexplained. In a direct laboratory experiment precipitating gypsum, Lloyd (1968) found h ,30 = 2.0°/0o. Further, " A number of pairs of gypsum and associated brines from marine evaporating pans have been examined. The average difference between the gypsum and dissolved sulfate is 3.60/00 (actually +3.6 + 0.9% 0: Lloyd, 1967, table I), which is in fair agreement with the experimental d a t a . . . However, there is no way of determining from this experiment whether this factor represents equilibrium fractionation or a kinetic fractionation." Sakai (1972) analyzed two modern marine gypsums and found A 130 = +2.3, +7.20/00. Although possible values for ~ 130 range from 2.0 to 9.5°/00, the best choice at present seems to be 3.6% o, as this represents a situation closest to the natural system of interest to us. Thus evaporites may be expected to be enriched in 180 by this a m o u n t with respect to the sulfate ion in the water from which they crystallized, and the removal of this heavy sulfate makes the residual sulfate correspondingly light. Putting all of this together, we propose that the oxygen in sulfate ions of present-day seawater is maintained at 5 ,O = + 8 . 6 ° o by a dynamic balance of: (a) an input of river water with 840 ~ + 5 ° o , and (b) an o u t p u t that tends toward a - 3.60o change due to sulfate precipitation into evaporites and at the same time tends toward a +10%0 change due to sulfate reduction.

Using these values of inputs and outputs we can make a material balance to calculate the mass fraction x = dM,2/(dM12 + dM,3) (Fig. 2) of the o u t p u t that must be contributed by sulfate reduction to maintain the sea (of constant sulfate mass) as high as ~ l = +8.6%o : 10x = 3.6(1 - x) + (8.6 - 5.0) x = 0.53


By a r o u n d a b o u t balance of the whole sulfur cycle, Holser and Kaplan (1966) calculated t h a t in Neogene times x = 0.36, so the above approximations seem reasonable. POSSIBLE ISOTOPIC E Q U I L I B R I U M B E T W E E N IN T H E S E A



In order for the proposed mechanism for the maintenance of 8 '80 in seawater sulfate to work, the sulfate ion must be effectively isolated from exchange of oxygen isotopes with water oxygen, over a long period of time. Longinelli and Craig (1967) assumed seawater to be at equilibrium with sulfate ion. Lloyd (1967, 1968) found a fractionation factor so high (A '80 = +38°/00 at low temperature, Table I) that this equilibrium was precluded. Lloyd also determined experimentally that a 97% exchange of oxygen between water and sulfate would require about 0.25 Ma. If correct, this exchange rate is slow enough to allow mixing of the oceans, and subsequent precipitation of sulfate from a batch of such equilibrated ocean water in an evaporite basin. But new data from interstitial waters in deep sea cores, that show no appreciable change in oxygen isotope ratio in the sulfate of interstitial waters going back to the late Cretaceous, suggest no appreciable exchange between sulfate ion and water for times much longer than the residence time of sulfate in the sea (20 Ma) (Zak et al., in prep.) This (unexplained) delay in equilibration leads to the result that 5 ~8Oso ' in the sea follows the dynamic balance of 5'8Oso" for inputs and outputs to the sea, rather than equilibrating with 8~8OH~o of seawater. POSSIBLE F A S T SEA-FLOOR REDUCTION--OXIDATION CYCLE

Lloyd (1967) proposed that the seawater sulfate oxygen isotope ratio was maintained by a relatively fast reduction--oxidation cycle at the sea--mud boundary, in which the H~S generated below the boundary is re-oxidized by dissolved air just above the boundary, and continuously re-cycled. However, his calculations should be modified, which makes his interpretation less likely. Firstly, Kroopnick and Craig (1976) have shown that ocean-bottom water is enriched by A180 = +3 to +7%0, relative to the starting value for air of 8180 = + 2 3 ° o assumed by Lloyd (1967). Secondly, we have given reasons above for believing t h a t the fractionation of oxygen attendant on sulfate reduction is more like A180 = -10% 0, than the experimental value

of - 4 . 7 0 o measured by Lloyd. Assuming b o t t o m water enrichment of /x180 = +5°/00, air at ~ ~80 = +230/00, and fractionations for air of A ~80 = -8.7°/oo and for sulfate reduction of A x80 = - 1 0 ° o , the predicted value for this system would be (Lloyd, 1967, eq. 4): 0.32(23 + 5 - 8.7) + 10 = 16.2°/00. This is so much higher than the 6 ~80 = 8.6%0 observed in present seawater that reduction--oxidation in the sea appears to be relatively ineffective in comparison with the long-term inputs and outputs shown in Fig. 2. As this paper was going to press, Fontes and Pierre (1978) presented some new analyses of 518Oso ' and 6 34Sso, in brines from evaporation ponds, that they interpret in terms of Lloyd's reduction--oxidation model, but for the above reasons (and others that cannot be discussed here) this process is probably not efficient on an oceanic scale. Finally, we point out that nearly all of the marine sulfate minerals of every age have 5 ~80 in the range +10 to +16%0 (Sakai, 1972; Claypool et al., 1972), This is so low that it indicates minor to negligible precipitation from, or re-equilibration with, sulfate having either the low-temperature equilibrium value of ~ 1sO = +380/00, or the sea-floor reduction--oxidation value of ~ 180 = +16%0. POSSIBLE POST-DEPOSITIONAL



In view of the generally held thesis that 5 ~8Oco 3 in old limestones has been substantially altered from its depositional value (e.g., Veizer and Hoefs, 1976), such alteration must also be considered for 5 lSOso ' in sulfate minerals. Two cases will be considered: equilibration of oxygen between sulfate ions of the mineral and water of dehydration, and equilibration of oxygen between sulfate of the mineral and an excess of subsurface water. The typical evaporite gypsum transforms to anhydrite as the temperature is raised on burial to the range 50--100°C, and conversely the anhydrite often alters to gypsum under the action of cold groundwater as the formation is exhumed (e.g., McDonald, 1953). At the higher temperatures, especially during initial burial, there should be maximum opportunity for the oxygen of the sulfate minerals to come to equilibrium with oxygen of any water that may be present. In particular, Heard and R u b e y (1966) have theorized that when gypsum dehydrates, the new anhydrite might retain the product water as pore water under pressure. We examine this case to see what the resulting equilibrium of ~ 1sO between crystal sulfate and liberated water should be. Gonfiantini and Fontes (1963) have measured a fractionation of A 180 = +4%0 for the crystallization of water of hydration of gypsum, relative to the solution water. Evaporation experiments by Gonfiantini (1965) indicate that seawater evaporated sufficiently to precipitate gypsum (about 25% remaining) should have raised ~ 180 to about +9%0; putting these two numbers together indicates that the water of crystallization should be about AIsO = + 1 3 ° o relative to ocean water. Gypsum crystallizing from an artificial salt pan gave

1sO = +12.3 + 0.4 (Fontes, 1966), from a sabkha bordering the Mediterranean gave 8 ~so = +7.8 + 2.6 (Sofer, 1975), and from Laguna Madre, Texas, gave ~t~80 = +3.2 (Matsubaya and Sakai, 1973). Assuming as a c o m m o n example that ~ ~8Oso 4 Cx) = +12°/oo and 8 IsOH20(X ) = +11%o, the total oxygen in the crystal (and in the gypsum rock if it has no porosity) is 5 laOt(x) = [(2 • 12) + + 11]/3 = 11.6. The data of Lloyd for the equilibration of anhydrite with water oxygen indicate a fractionation A ISOso _H2 o = + 24.7°/00 at 100°C (Table I, column 1); balance b e t w e e n oxygen of new anhydrite and the released water of crystallization (half as much oxygen) gives, for example: [2 • 5 'SOso,(x) + ~ I a O H ~ o ( 1 ) ] / 3 = (~ lsO(t ) 3 • ~ ~8OsoAx) -- h ~ S O s o _ H ~ o = 3 " ~ 1 8 0 ( t ) 3 • 518Oso4(x)


24.7 = 3 • 11.6

lSOso,(x) = +19.8%0 ~ +20%0; ~ 18OH20(1 ) = - 4 . 9 % 0 ~ - 5 % 0


Analogous calculations for other temperatures give the results in column (6) of Table I. At temperatures where water is likely to be released from gypsum, in the range 50--150°C, these hypothetical values of ~ ~sO are substantially higher than those actually measured in evaporite rocks of any age (Sakai, 1972; Claypool et al., in prep.). Subsurface waters have a wide range of 5~soH: o. Clayton et al. (1966) have explained many of these as the result of equilibration at various temperatures (in the range 100--300°C) with limestone that has been brought d o w n from its initial value of 5 XSOcaco3 = +300/00 (SMOW), to a c o m m o n l y observed value of around +24%0. If anhydrite were similarly altered b y exchange or recrystaltization with the oxygen of subsurface waters at moderate temperatures, the resulting 5 ~SOcaso 4 would depend on the temperature, the relative proportions of anhydrite and water, and the degree of approach to equilibrium. As an example, we assume simultaneous equilibrium of h o t subsurface water with a large a m o u n t of calcite and a small a m o u n t of anhydrite [somewhat on the model of Clayton et al. ( 1 9 6 6 ) ] , which is analogous to equilibrium between calcite and anhydrite. Equilibrium data from Lloyd (1968) for anhydrite-water (Table I, column 1) and from O'Nefl et al. (1969) for calcite-water (Table I, column 4), give, for example, at 200°C: h 'SOcaso _CaCO" = 14.0 -- 9.1 = +4.90/00. So if the limestone has ~ '8Ocac% = 240/00, one would predict '8Ocaso" ~ +28.9°/o0 . Calculations at other temperatures are given in column 7 of Table I. Evidently all post-depositional alterations of anhydrite would tend to raise 18Ocaso * into the range +20 to +300]00. The measured values on old anhydrite are generally in the range 51sO = +10 to +16°/0o (Sakai, 1972; Claypool et al., 1972, and in prep.). Consequently, one can say that while some of the dispersion of 8 x80 is possibly caused b y a slight adjustment of the anhydrites toward either of these equilibria with water, certainly there has been no general degradation of ~ 18Ocaso, in the way that there has been for ~ lSOcac % .

10 ISOTOPIC EQUILIBRATION OF SULFATE AND WATER OXYGEN DURING HYDROTHERMAL CIRCULATION OF SEAWATER Recent discussions have emphasized the importance of a hydrothermal circulation of seawater at mid-ocean ridges for chemical interactions of seawater and basaltic rocks (Spooner and Fyfe, 1973; Wolery and Sleep, 1976). In particular, isotopic equilibration between water and rock in the hydrothermal system has been postulated for 180/'60 (Muehlenbachs and Clayton, 1976; O h m o t o et al., 1976) and for S~Sr/S6Sr (Spooner, 1976). The question of interest here is whether 518Oso ' of the ocean as a whole can be, or has been, substantially affected by such hydrothermal activity. This is a very complex question, and here we will attempt no more than a summary of the evidence. Three processes may affect concentration or isotopic (534S and ~ 1sO) ratios of the sulfate ion: (1) sulfate may be removed--precipitated in the subsea basalt as anhydrite as a result of both its retrograde solubility with rising temperature, and the rise of calcium concentration by leaching from the basalt; (2) sulfate may be reduced to pyrrhotite, pyrite, or sulfide (H2S) by oxidation of ferrous iron in the sequence fayalite--magnetite--hematite, splitting off the oxygen of the sulfate into magnetite, hematite or chlorite; (3) either in conjunction with or independently of the above reactions, ~ 18Oso ' in any remaining sulfate may equilibrate with ~ lsOH2 o of the h o t water or with 61~O silicate of the hot rock. These possibilities are considered in the following paragraphs. Precipitation of anhydrite has been demonstrated in laboratory experiments (Bischoff and Dickson, 1975; Hajash, 1975; Mottl et al., 1974; Mottl and Seyfried, 1977; Scott and Frank, 1974) and has been found in at least one Recent marine hydrothermal circulation system (Bjornsson et al., 1972) and in one non-marine system (McDowell and McMurry, 1977). But the experiments were quenched, and even during quenching anhydrite had begun to redissolve (Bischoff and Dickson, 1975; Hajash, 1977); consequently one might expect that during the cooler waning phase of a natural hydrothermal system {Lowell, 1975} some of the anhydrite that had been precipitated at higher temperatures would be re-dissolved and returned to the sea. Theoretical calculations by Reed (1977) come to the same conclusion. Anhydrite is not present in metamorphosed oceanic basalts (Bischoff and Dickson, 1975; Hajash, 1975), and it is also missing in stratiform sulfide deposits that have developed from oceanic hydrothermal systems (e.g., Large, 1977). In Kuroko-type stratiform deposits the stratigraphic and isotopic data indicate that the anhydrite/gypsum which is a prominent c o m p o n e n t of such deposits originated directly from the seawater overlying a hydrothermal vent and not from the hydrothermal solution itself. From such evidence, it is hard to see h o w the removal of sulfate as anhydrite, b y circulation of seawater in ocean-ridge hydrothermal systems, can have had any quantitative effect on sulfate concentration in seawater or isotope ratios in sulfate. But if the circulation of seawater through such hydrothermal systems is as massive (Spooner and Fyfe, 1973; Wolery and Sleep, 1976) as

11 seems to be required by heat flow (e.g., Williams et al., 1974), it is also hard to believe that the large amount of anhydrite, that would have had to be deposited at moderate temperatures, could have been totally removed by later circulation, especially as the circulation flux decreases sharply as the temperature drops (Lowell, 1975). Laboratory experiments, particularly those of Hajash (1975) and Hajash and Tieh (1976) emphasize the appearance of sulfide minerals (pyrrhotite, chalcopyrite) during the interaction of seawater with basalt at high temperatures. In conjunction with thermodynamic calculations that show sulfide predominant over sulfate at these conditions, it has generally been concluded (with varying degrees of qualification) that such sulfides could have resulted from reduction of seawater sulfate (Spooner and Fyfe, 1973; Scott and Frank, 1974; Bischoff and Dickson, 1975; Hajash, 1975; Wolery and Sleep, 1976; Bonatti et al., 1976). But most of the experimental results so far available are qualitative, and other evidence appears to show that sulfur may move from rock to solution rather than from solution to rock. Oceanic basalts erupt saturated with magmatic sulfur (Moore and Fabbi, 1971; Mathez, 1976) and during greenstone metamorphism (by hydrothermal circulation of seawater) may lose as much as 90% of their sulfur (Scott and Frank, 1974; Peron and Scott, 1976). Grinenko et al. (1975) analyzed altered and unaltered mid-ocean ridge basalts, but the sulfur contents are not significantly different. Spooner et al. (1977a) find no secondary sulfides in one example of a sequence of altered oceanic pillow lavas. Furthermore, hydrothermally altered oceanic basalts do not show the large change of ferrous/ferric ratio expected from reduction of sulfate to sulfide (Humphris and Thompson, 1978). So, while it may be possible that oceanic seawater circulation systems are responsible for the local deposition of certain sulfide ores, it is difficult to support a case for large-scale reduction that would affect the concentration or isotopic ratio of sulfate in the whole ocean. Another reason to be skeptical about large-scale removal of oceanic sulfate in hydrothermai systems, as either anhydrite or as sulfides, is the present relatively high level of sulfate concentration in the sea. Wolery and Sleep (1976) estimate that the heat flow data require a circulation of over 1017 g/y of seawater through oceanic hydrothermal systems. That would cycle the whole mass of the ocean ("hydrothermal recycling time" of Spooner, 1976) in 3 to 10 Ma compared to the residence time of 22 Ma for sulfate in the sea calculated from long-term river input (Holser and Kaplan, 1966). If indeed sulfate is substantially removed during hydrothermal circulation, as either anhydrite or sulfides, we would expect that the hydrothermal cycle would control sulfate in the sea at a much lower level than we now find it. For example, the hydrothermal system is estimated to raise the temperature of circulated seawater to a maximum temperature estimated at 200--400°C (Spooner and Fyfe, 1973; Wolery and Sleep, 1976). Solubility of anhydrite, decreasing with an increase in temperature, begins to precipitate anhydrite at 135°C, and reduces its concentration in seawater to a half at 160°C and a tenth at 260°C (Bischoff and Dickson, 1975; data of Blount and Dickson, 1969). Reduction to sulfide

12 and equilibration with pyrite--pyrrhotite would similarly reduce total sulfur b y several orders of magnitude, particularly at temperatures below 300°C (Kajiwara, 1976), so that even if the sulfide was re-oxidized when it re-entered the sea, the level of sulfate concentration in the sea would be controlled at some very low level. If hydrothermal circulation of seawater is actually as massive as implied by the heat-flow data, then we have to consider the possibility of equilibration at high temperatures between sulfate oxygen and water oxygen. Consider the length of time during which the seawater has an opportunity to equilibrate at high temperature; this is somewhat less than the residence time of seawater in the hydrothermal system. The residence time: Tw = (Ar

• hh

" th " P

" Pw



where Ar is the rate of formation of area of new oceanic basalt per year, estimated at 3 • 10 is cm 2 (Williams and Von Herzen, 1974); hh is the depth of the hydrothermal circulation in fractured basalt (variously estimated at 10 s cm (Muelenbachs and Clayton, 1976), 4.5 • l 0 s cm (Spooner, 1976), and 5 • l 0 s cm; Wolery and Sleep, 1976); th is the duration of the hydrothermal system, variously estimated at 1.7--2.3 • 107 y (Wolery and Sleep, 1976), 5-80 • 106 y (Langseth et al., 1977), 3 • l 0 s y (Scott et al., 1976), and perhaps 104 y (Lowell, 1975); p is the porosity of the basalt, estimated at 0.1--1% (Wolery and Sleep, 1976); Pw is the density of seawater; and Fw is the flux of seawater through the basalt, estimated at 1.3 to 9.101~ g/y (Wolery and Sleep, 1976) and 4.5.1017 g/y (Spooner, 1976). So depending on what numbers y o u believe, a residence time for seawater can. be calculated from less than one to more than a thousand years (Wolery and Sleep, 1976, calculated 100--10,000 y). Comparison with the measured half-time for equilibration, of oxygen isotopes between sulfate ion in solution and water molecules, of about 10 y at 200°C (Lloyd, 1968), implies that substantial oxygen exchange of this type could have occurred, b u t was n o t necessarily complete, Assuming equilibrium was attained, with seawater " b u f f e r e d " near ~ ISOH~o = 0 (Muehlenbachs and Clayton, 1976), Lloyd's (1968) measured fractionation gives, for example at 200°C, 51SOso,(1)= +9.0%0 in whatever part of the sulfate remained in solution at that temperature. But this is probably, as explained above, only a small fraction of the seawater sulfate. Incidentally, n o t only calcium b u t also strontium is also removed with the sulfate (a factor that may be important for strontium isotope geochemistry). If most of the seawater sulfate is removed as anhydrite, and later re-dissolved at low temperature and returned to the sea, t80 in this major fraction will be virtually unchanged by the process. Alternativeiy, if the circulating seawater sulfate is reduced to sulfide, and later reoxidized on the sea floor, this overall process will substitute heavy atmospheric oxygen like that calculated above for the hypothetical fast sea-floor oxidation scheme. Isotope equilibrations in the sea-floor hydrothermal system have previously been postulated for laO/160 between water and silicates (Muehlenbachs and

13 Clayton, 1976; Spooner et al., 1977a), and for 87Sr/86Sr between Sr in solution and in rock (Spooner, 1976; Spooner et al., 1977b). Despite the optimism of its proponents, data supporting isotope exchange in basalts and the interacting seawater are still somewhat equivocal. Muehlenbachs and Clayton (1976) find significant a depletion in 8180 of 1.3°/00 in hydrothermally altered whole rocks, but their table 2 shows that this is more vaguely 1.5 + 1.2°/00. Such a decrease in 8180 is supported by the experiments of O h m o t o et al. (1976). Spooner et al. (1977a) find, on the contrary, a substantial increase of 8 '80 in altered basalts, and their calculations do not seem to explain this reversal of sign. Spooner (1976) and Spooner et al. (1977b) found an increase of 8~Sr/86Sr of 0.0013 + 0.0009 in an altered series of basalts, and initially interpreted this in terms of a corresponding decrease in 87Sr/86Sr of hydrothermally circulated seawater (Spooner, 1976). But a later m o r e detailed t r e a t m e n t ' o f the problem concluded that the circulating seawater is so inefficient in exchange that it should show no appreciable change in 87Sr/86Sr (Spooner et al., 1977b). All of these considerations indicate some of the difficulties of generating any substantial change in either the concentration or isotope level (for either oxygen or sulfur) of sulfate in the sea, by massive hydrothermal circulation beneath mid-ocean ridges. This conclusion m a y be difficult to reconcile with the evidence for massive circulation of seawater. CONCLUSIONS AND QUESTIONS The sulfate ion seems to be a very stable unit, which allows little change in its isotope ratios of oxygen (or sulfur) once the ion is formed. Analyses of deep-sea cores indicate a rate even slower than that given by experiment, of isotope exchange of oxygen between sulfate and water, and analyses of evaporite samples indicate that 8 'sO in sulfate minerals has not been equilibrated by contact with either released water of hydration or with formation waters. Also, the higher temperatures of a subsea hydrothermal circulation system, where equilibration of sulfate with water would be greatly accelerated, seems to be ineffective in altering 8 'sO of the sulfate of the sea as a whole. Consequently, 8 '80 in sulfate minerals of evaporites is a sample of that in sulfate of contemporary seawater, which in turn is controlled by a dynamic balance of the results of sedimentation outputs of sulfur as sulfide m u d ( 4 , 2 0 ~ +10°/o0) and sulfate evaporite (4130 ~ -3.60/00) and the erosion inputs from the weathering of sulfides (82,0 ~ 20/00) and sulfates (8310 ~ +12°/00). With the exception of the weathering of sulfide, these fluxes are the same as those that control the isotope ratio of sulfur. These conclusions, however, are only provisional. Many questions are still unresolved: (1) Why does 8 '8Oso, remain constant in deep-sea cores, unreacted with 8 'sOH20, for times much longer than indicated by Lloyd's laboratory experiments?


(2) Is the best mean value of the ratio of sulfur to oxygen fractionation during biological reduction of sulfate about 4, or is it substantially less? (3) Will field studies of pyrite oxidation, including isotope measurements on the oxygen of the air and the water, ratify the ratio of 1:2 of those participants in the oxidation, as measured in Lloyd's experiment? (4) What was the long-term, pre-industrial mix of weathered sulfide and evaporite sulfate in river input to the sea, and what were its isotope ratios (of O and S)? (5) If, as claimed, a substantial mass of seawater has circulated in hydrothermal systems at mid-ocean ridges, then what has happened to the seawater sulfate in this process -- its form, concentration and isotope ratios (of O and S)? ACKNOWLEDGEMENTS

This research was begun by the group working at the University of California, Los Angeles, in part supported b y the U.S. National Aeronautics and Space Administration Grant No. NGR05-007-221. Holser continued in the Abteilung fiir Chemie der A t m o s p h ~ e , Max-Planck-Institut fi]r Chemie (Otto-HahnInstitut), Mainz, Germany, under a Senior U.S. Scientist Award of the Alexander von Humboldt-Stiftung; and in the Department of Materials Science, Tohoku University, Sendai, Japan, under the U.S.--Japan Cooperative Science Program, U.S. National Science Foundation Grant INT 76-02721.

REFERENCES Berner, R.A., 1971. Worldwide sulfur pollution of rivers. J. Geophys. Res., 76: 6597--6600. Bischoff, J.L. and Dickson, F.W., 1975. Seawater-basalt interaction at 200°C and 500 bars: implications for origin of sea-floor heavy-metal deposits and regulation of seawater chemistry. Earth Planet. Sci. Lett., 25: 385--397. Bjornsaon, S., Aronoson, S. and Tomasson, J., 1972. Economic evaluation of Reykjanes thermal brine area, Iceland. Am. Assoc. Pet. Geol. Bull., 56: 2380--2391. Blount, C.W. and Dickson, F.W., 1969. The solubility of anhydrite in NaCl-H20 from 100 to 450°C and 1 to 1000 bars. Geochim. Cosmochim. Acta, 33: 227--245. Bonatti, E., Guerstein-Honnorez, B.M. and Honnorez, J., 1976. Copper--iron sulfide mineralizations from the equatorial Mid-Atlantic Ridge. Econ. Geol., 71: 1515--1525. Chase, C.G. and Perry, E.C., Jr., 1972. The oceans: growth and oxygen isotope evolution. Science, 177: 992--994. Claypool, G.C., Holser, W.T., Kaplan, I.R., Sakai, H. and Zak, I., 1972. Sulfur and oxygen isotope geochemistry of evaporite sulfates (abstr.). Geol. Soc. Am. Abstr. Progr., 4: 473. Claypool, G.C., Holser, W.T., Kaplan, I.R., Sakai, H. and Zak, I., in prep. The age curves of sulfur and oxygen isotopes in marine sulfate and their mutual interpretation. Clayton, I.F., Graf, D.L., Mayeda, T.K., Meents, W:F. and Shimp, N.F., 1966. The origin of saline formation waters. 1. Isotopic composition. J. Geophys. Res., 71: 3869--3882. Emiliani, C. and Shackleton, N.N., 1974. The Brunhes Epoch: isotopic paleotemperatures and geochronology. Science, 183:511--514. Fontes, J.C., 1966. Int~r~t en g~ologie d'une dtude isotopique de l'dvaporation. C. R. Acad. Sci., Paris, Sgr. D, 1950--1953. Fontes, J.-C. and Pierre, C., 1978. Oxygen 18 changes in dissolved sulfate during sea water evaporation in saline ponds (abstr.). Int. Congr. Sedimentology, 10th, Abstr., 1: 215--216.


Garlick, G., 1969. Oxygen. Isotope Variations in Nature. In: K.H. Wedepohl (Editor), Handbook of Geochemistry. Springer, Berlin, chap. 8B. Garrets, R.M., MacKenzie, F.T. and Hunt, C., 1973. Chemical Cycles and the Global Environment. William Kaufmann, Inc., Los Altos, California, pp. 206. Goldhaber, M.B. and Kaplan, I.R., 1974. The sulfur cycle. In: E.D. Goldberg (Editor), The Sea, Vol. 5. Wiley, N e w York, N.Y., pp. 569--655. Gonfiantini, R., 1965. Effeti isotopi nell'evaporazione di acque salate. Atti Soc. Toscana Sci. Nat. Pisa, Mere., P.V., Ser. A, 72: 1--22. Gonfiantini, R. and Fontes, J.C., 1963. Oxygen isotopic fractionation in the water of crystallization of gypsum. Nature, 200: 644--646. Grinenko, V.A., Migdisov, A.A. and Barskaya, N.V., 1973. Sulfur isotopes in the sedimentary mantle of the Russian Platform. Dokl. Akad. Nauk. SSSR, 210:445--448 (transl. Dokl. Earth Sci. Sec., 210: 224--227, 1973.) Grinenko, V.A., Dimitriev, L.V., Migdisov, A.A. and Sharashkin, A.Ya., 1975. Sulfur contents and isotope compositions for igneous and metamorphic rocks from mid-ocean ridges. Geokhimiya, 199--206 (transl. Geochem. Int., 132--137, 1975). Hajash, A., 1975. Hydrothermal processes along mid-ocean ridges: an experimental investigation. Contrib. Mineral. Petrol., 53: 205--206. Hajash, A., 1977. Experimental seawater/basalt interactions: effects of water/rock ratio and temperature gradient (abstr.).Geol. Soc. Am. Abstr. Progr., 9: 1002. Hajash, A. and Tieh, T.T., 1976. Experimental seawater/basalt reactions, 200--500°C, 500--2000 bars: additional mineralogical data (abstr.).Geol. Soc. Am. Abstr. Progr., 8: 8 9 9 - 9 0 0 . Hartmann, M. and Nielsen, H., 1969.634S-Werte in rezenten Meeressedimenten und ihre Deutung am Beispiel einiger Sedimentprofile aus westlichen Ostsee. Geol. Rundsch., 58: 621--655. Heard, H.C. and Rubey, W.W., 1966. Tectonic implications of gypsum hydration. Geol. Soc. Am. Bull., 77: 741--760. Hoefs, J., 1973. Stable Isotope Geochemistry. Springer, New York, N.Y., 140 pp. Holser, W.T. and Kaplan, I.R., 1966. Isotope geochemistry of sedimentary sulfates. Chem. Geol., 1: 93--135. Humphris, S.E. and Thompson, G., 1978. Hydrothermal alteration o f oceanic basalts by seawater. Geochim. Cosmochim. Acta, 42: 107--125. Jurlo, R.L., 1976. Estudio de la DistributiSn de los Isotopos del Azufre (34S/3~S) en Sedimentos Pertenecientes Principalmente al Geosinclinal Variscio Europeo. Dissertation, Universidad Nacional de la Plata, Argentina. Kajiwara, Y., 1976. Ultimate origin of "ore solutions": ocean water? Geochem. J., 9: 235--239. Kaplan, I.R., Emery, K.O. and Rittenberg, S.C., 1963. The distribution and isotopic abundance of sulphur in Recent marine sediments off southern California. Geochim. Cosmochim. Acta, 27: 297--331. Kellog, W.W., Cadle, R.D., Allen, E.R., Lazrus, A.L. and Martell, E.A., 1972. The sulfur cycle. Science, 175: 587--596. Knauth, L.P. and Epstein, S., 1976. Hydrogen and oxygen isotopes in nodular and bedded cherts. Geochim. Cosmochim. Acta, 40: 1095--1108. Kroopnick, P. and Craig, H., 1976. Oxygen isotope fractionation in dissolved oxygen in the deep sea. Earth Planet. Sci. Lett., 32: 375--388. Langseth, M.B., Anderson, R.N. and Hobart, M.A., 1977. Seafloor heat flow measurements and the extent of water circulation in oceanic crust (abstr.). Geol. Soc. Am., Abstr. Progr., 9: 1065. Large, R.R., 1977. Chemical evolution and zonation o f massive sulfide deposits in volcanic terrains. Econ. Geol., 72: 5 4 9 - 5 7 2 . Lloyd, R.M., 1967. Oxygen ,s composition of oceanic sulfate. Science, 156: 1228--1231.

16 Lloyd, R.M., 1968. Oxygen isotope behavior in the sulfate-water system. J. Geophys. Res., 73: 6099--6110. Longinelli, A. and Cortecci, G., 1970. Isotopic abundance of oxygen and sulfur in sulfate ions from river water. Earth Planet. Sci. Lett., 7: 376--380. Longinelli, A. and Craig, H., 1967. Oxygen-18 variations in sulfate ions in seawater and saline lakes. Science, 156: 56--59. Lowell, R.P., 1975. Circulation in fractures, hot springs, and convective heat transport on mid-ocean ridges. Geophys. J. R. Astron. Soc., 40: 351--365. Mathez, E.A., 1976. Sulfur solubility and magmatic sulfides in submarine basalt glass. J. Geophys. Res., 81: 4269--4275. Matsubaya, O. and Sakai, H., 1973. Oxygen and hydrogen isotope study on the water of crystallization of gypsum from the Kuroko type mineralization. Geochem. J., 1 : 153-165. McDonald, G.J.F., 1953. Anhydrite--gypsum equilibrium relations. Am. J. Sci., 251: 884-898. McDowell, D. and McMurry, M., 1977. Active metamorphism in the Salton Sea geothermal field, California: mineralogical and mineral chemical changes with depth and temperature in sandstone (abstr.). Geol. Soc. Am., Abstr. Progr., 9: 1088. Migdisov, A.A., Cherkovskiy, S.L. and Grinenko, V.A., 1974. The effects of formation conditions on the sulfur isotopes of aquatic sediments. Geokhimiia, 1482--1502 (transl. Geochem. Int., 11: 1028--1045, 1974). Mizutani, Y., 1973. Isotopic behaviour of sulphate oxygen in the bacterial reduction of sulphate. Geochem. J., 6: 183--191. Mizutani, Y. and Rafter, T.A., 1969. Oxygen isotopic composition of sulphates -- Part 4. Bacterial fractionation of oxygen isotopes in the reduction of sulphate and in the oxydation of sulfur. N.Z.J.Sci., 12: 60--68. Moore, J.O. and Fabbi, B.P., 1971. An estimate of the juvenile sulfur content of basalt. Contrib. Mineral. Petrol., 33: 118--127. Mottl, M.J. and Seyfried, W.E., 1977. Experimental basalt-seawater interaction: rock-vs. seawater-dominated systems and the origin of submarine hydrothermal deposits (abstr.). Geol. Soc. Am. Abstr. Progr., 2: 1104. Mottl, M.J., Corr, R.F. and Holland, H.D., 1974. Chemical exchange between seawater and mid-ocean ridge basalt during hydrothermal alteration: an experimental study (abstr.). Geol. Soc. Am. Abstr. Progr., 6: 879. Mueblenbachs; K. and Clayton, R.N., 1976. Oxygen isotope composition of the oceanic crust and its bearing on seawater. J. Geophys. Res., 81: 4365--4369. Nakai, N., 1963. Biochemical oxidation of sulfur and sulfide minerals by mixed cultures, and the behavior of stable sulfur isotopes. J. Earth Sci. Nagoya Univ., 11: 279--296. Ohmoto, H., Cole, D.R. and Mottl, M.J., 1976. Experimental basalt-seawater interaction: sulfur and oxygen isotope studies (abstr.). EOS, 57: 342. O'Neil, J.R., Clayton, R.N. and Mayeda, T.K., 1969. Oxygen isotope fractionation in divalent metal carbonates. J. Chem. Phys., 51: 5547--5558. Peron, P.R. and Scott, R.B., 1976. Chemical character of hydrothermal alteration and weathering of oceanic basalts, Mid-Atlantic Ridge (abstr.). Geol. Soe. Am. Abstr. Progr., 8: 1074. Perry, E.C. and Tan, F.C., 1972. Significance of oxygen and carbon isotope determinations in early Precambrian cherts and carbonate rocks of Southern Africa. Geol. Soc. Am. Bull., 83: 647--664. Reed, M.H., 1977. Calculations of Hydrothermal Metasomatism and Ore Deposition in Submarine Volcanic Rocks with Special Reference to the West Shasta District, California. Ph.D. Thesis, University of California, Berkeley, Calif. Rees, C.E., 1973. A steady-state model for sulfur isotope fractionation in bacterial reduction processes. Geochim. Cosmochim. Acta, 37: 1141--1162. Sakai, H., 1972. Oxygen isotope ratios of some evaporites from Precambrian to Recent ages. Earth Planet. Sei., 15: 201--205.

17 Sakai, H. and Matsubaya, O., 1974. Isotopic geochemistry of the thermal waters of Japan and its bearing on the Kuroko ore solutions. Econ. Geol., 69: 974--991. Schwarcz, H.P. and Burnie, S.W., 1963. Influence of sedimentary environments on sulfur isotope ratios in clasticrocks, a review. Miner. Deposita, 8, 264--277. Schwarcz, H.P. and Cortecci, G., 1974. Isotopic analyses of spring and stream water sulfate from the Italian Alps and Appennines. Chem. Geol., 13: 285--294. Scott, R.B. and Frank, D.J., 1974. Distribution of sulfur in the ocean crust (abstr.).Geol. Soc. Am. Abstr. Progr., 6: 945. Scott, R.B., Malpas, J., Rona, P.A. and Udintsev, G., 1976. Duration of hydrothermal activity at an ocean spreading center, Mid-Atlantic Ridge (lat.26°N). Geology, 4: 233-236. Smejkal, V., 1978. Oxygen isotopic composition of sulphates from some mineral waters and mine waters in western Bohemia. Int. Symp. on Isotope Hydrology, Neuherberg, Jun. 19--23, 1978. Sofer, Z., 1975. Isotopic Composition of Hydration Water of Gypsum. Thesis, Weizmann Inst. of Science, Rehovot, Israel(in Hebrew, English summary). Spooner, E.T.C., 1976. The strontium isotopic composition of seawater, and seawateroceanic crust interaction. Earth Planet. Sci. Lett., 31: 167--174. Spooner, E.T.C. and Fyfe, W.S., 1973. Sub-sea-floor metamorphism, heat, and mass transfer. Contrib. Mineral. Petrol., 42: 287--304. Spooner, E.T.C., Beckinsale, R.D., England, P.C. and Senior, A., 1977a. Hydration, '80 enrichment and oxidation during ocean floor hydrothermal metamorphism of ophiolitic metabasic rocks from E. Liguria, Italy.Geochim. Cosmochim. Acta, 41: 857--871. Spooner, E.T.C., Chapman, H.J. and Sinewing, J.D., 1977b. Strontium isotopic contamination and oxidation during ocean floor hydrothermal metamorphism of the ophiolitic rocks of the Troodos Massif, Cyprus. Geochim. Cosmochim. Acta, 41: 873--890. Sweeney, R.W. and Kaplan, I.R., in prep. Diagenetic sulfate reduction in marine sediments. (Submitted to Marine Chem. ) Sweeney, R.W. and Kaplan, I.R., in prep. Isotope fractionation of sulfur during sulfate reduction in marine sediments. (Submitted to Marine Chem.) Veizer, J. and Hoefs, J., 1976. The nature of '*0/~0 and ~3C/~2C secular trends in sedimentary carbonate rocks. Geochim. Cosmochim. Acta, 40: 1387--1395. Wagener, K., 1973. Entwicklung der irdischen Atmosph~ire durch Evolution der Biosph~ire. Rheinisch-Westf~il. Akad. Wiss., Natur., Ing. Wirtschaftswiss.,Vortr., 233: 36. Williams, D.L. and V o n Herzen, R.P., 1974. Heat loss from the earth: new estimate. Geology, 2: 327--328. Williams, D.L., V o n Herzen, R.P., Sclater,J.G. and Anderson, R.N., 1974. The Galapagos spreading center: lithosphericcooling and hydrothermal circulation. R. Astron. Soc. Geophys.J., 38: 587--608. Wolery, T.J. and Sleep, N.H., 1976. Hydrothermal circulation and geochemical flux at midocean ridges. J. Geol., 84: 249--275. Zak, I., Sakai, H. and Kaplan, I.R., in prep. Isotopic compositions of sulfates in interstitialwaters from deep-sea sediments, and factors controlling the oxygen isotopic ratios of ocean sulfates.