9402693 EDWARDS Direct dating of the marine S18 O record has been a classic problem for the last four decades, and is still largely unsolved. Interest in this problem has increased in the wake of improvements in analytical capabilities for measurement of 234 U/ 238 U (Chen et al., 1986) and 230 Th/238 U (Edwards et al., 1987 a,b) and further analytical refinements (Edwards et al., 1993). However, attempts to use 230 Th techniques to date deep-sea marine carbonates will be faced with at least one major, perhaps insurmountable, obstacle: separation of contaminant 230 Th produced by uranium in the overlying water column from 230 Th produced by uranium in the carbonate (Delaney and Boyle, 1983). Consequently, two alternate approaches will be investigated to constrain the timing of the marine S18 O record: 234 U/238 U dating of carbonate-rich sediments (Ku, 1965; Ludwig et al., 1991) and 230 Th dating of tropical-island speleothems. The 234U/238U ratio is measured with a precision of +1 per mil (Edwards et al., 1993), which corresponds to an error in 234U/238U age of +4Ka. Thus high-resolution 234U/238U ages are in principle, possible. Bulk carbonate will be analyzed from cores that have high primary uranium contents and are close to 100% carbonate. Also, dating of speleothems from tropical islands will be dated with TIMS 230 Th techniques (Li et al., 1989; Dorale et al., 1992; Baker et al., 1993; Richards et al., in press) and may have S18 O values that mimic marine values, thereby providing a means of dating the marine record. These two approaches will place direct constraints on the timing of the marine S18 O record (over the past 350,000 years) and will help to understand the fundamental causes of Quaternary climate change.
