20. C, Emiliani, Bull, Geol. Soc. Amer. 75, 129 (1964). 21. W. L. Donn, W. R. Farrand, M. Ewing, J. Geol, 70, 206 (1964). 22. A. S. Merrill, K. O. Emery, Meyer Rubin, Sctence 147. 398 (1965). 23. J. C. Medcof, A. H. Clarke, Jr., J. S. Erskine, J. Fish, Res. Bd. Canada 22 (2), 631 (1965). 24. J. R. Curray, in The Quaternary of the United States, H. E. Wright and D, G. Frey, Eds. (Princeton Univ. Press, Princeton, NJ, 1965), p, 723. 25. F. C. Whitmore, Jr,, K. O, Emery, H. B. S. Cooke, D. J. P. Swift, Science 156, 1477 (1967). : 26. K. O. Emery and R. L. Edwards, Amer. Antiquity 31, 733 (1966). 27. W. S,. Newman, unpublished thesis, New York Univ. (1966); W. S. Newman and R. W. Fairbridge, in Proc. First Nat. Coastal and Shallow Water Res. Conf. (National Science Foundation and Office of WNavaf Research, Tallahassee, Florida, 1962, p. 188. 28. W. Harrison, R. J. Malloy, G. A. Rusnak, J. Terasmae, J. Geol. 73, 201 (1965). 29. K. QO. Emery, in preparation. 30. M. B. Davis and E, 5. Deevey, Science 145, 1293 (1964). 31. W. F. Libby, Radiocarbon Dating (Univ. of Chicago Press, Chicago, ed. 2, 1955). 32. P. Butler, Ecology 40, 735 (1959), 33. C. A. Kaye and E. S. Barghoorn, Bull. Geol. Soc. Amer. 75, 63 (1964), 34. G. W. Barendsen, E. 8S. Deevey, L. J. ‘35. 36. 37. 38. 39. Gralenski, Secfence 126, 908 (1957); E. S. Deevey, L. J. Gralenski, V. Hoffren, Amer. J. Set, Radiocarbon Suppl. 1, 144 (1959); Deevey (personal communication, April 1967) now accepts the date 15,090 years before the present for Totoket Bog. C. R. Groot and J. J. Groot, Amer. J. Sci. 262, 488 (1964), E. S. Deevey, IJr., Amer. J. Sci, 246, 329 (1948), W. S. Benninghoff, Phillips Acad. Peabody Foundation, Archaeol, Papers 2, 96 (1942), E. A. Stanley, Deep-Sea Res. 13, 921 (1966). J. G. Ogden, II], Amer. J. Sci. 261, 344 (1963). 40. E. B. Leopold, Prec. Nat. Acad. Sci. U.S. §2, 863 (1956). 41. E. Uchupi, Misc. Geol. Invest. Map I-451 (U.S. Geological Survey, Washington, D.C., 1965), 42. M. B. Davis, in The Quaternary of the United States, H. E. Wright and D. G. Frey, Eds. (Princeton Univ. Press, Princeton, NJ., 1965), p. 377. 43, Appreciation is due Captain Norman Lepire of New Bedford, Massachusetts, for Several sampies of peat and oyster shells, and to the U.S. Geological Survey through its support by contract 8358 to the Woods Hole Oceanographic Institution, Contribution No, 1968 of the Woods Hole Oceanographic Institution. Publication approved by the director, U.S. Geological Survey. 23 June 1967 a liter/min for 90 minutes. The condensable gases (including radon) were continuously collected in. traps cooled with liquid air, The radon was thenseparated from CO. and H:0 by circulating the condensed gas through ascarite. Next it was quantitatively transferred to a 30-cmscintillation cell where the count of its a-particles was determined. Overall recovery yields averaged 90 percent, accuracy of the measurements is about +10 percent. Details of the procedure have already been published (/). Although our results change neither the average content of radium content in the ocean nor the broad picture of its distribution from that given by previous workers (2, 3}, they do show far less scatter and suggest that radium is uniformly mixed throughout a given water mass. The radium content of surface water in both the Atlantic and Pacific oceans is 4 X 10-'* g/liter (Table 1 and Fig. 1). In both oceans a Radium-226 and Radon-222: Concentration in the Abstract. Measurements of radon-222 in seawater suggest the following. The radium-226 content of surface water in both the Atlantic and Pacific oceans is uniformly close to about 4 * 10™" pram per liter. The deep Pacific has a concentration of radium-226 that is four times higher and the deep Atlantic a concentration twice as high as that of the surface. These distribution profiles can be explained by the same particle-settling rate for radium-226 from surface to depth for the two oceans and by a threefold longer residence time of water in the deep Pacific than in the deep Atlantic. The vertical distribution of the deficiency of radon-222 in the surface water of the northwest Pacific Ocean suggests a coefficient of vertical eddy diffusion as high as 120 square centimeters per second and a gas-exchange rate for carbon dioxide in surface water between 14 and 60 moles per square meter per year. Vertical profiles of the excess of radon-222 in nearbottom water of the South Atlantic give coefficients of vertical eddy diffusion ranging from 1.5 to more than 50 square centimeters per second. change rates of gases across the air-sea interface, and (iii) the rates of vertical mixing near the surface and near the bottom of the ocean. Briefly, Rn*?? (half-life, 3.85 days) is produced in seawater by the decay of its parent Ra2** (half-life, 1600 years). Well away from the air-sea and sediment-sea inter- faces, the rate of radioactive decay of radon is equal to that of its parent radium. Thus, a measurement of the radon content of such a water sample is a measurement of its radium content. As the radon content of the atmosphere &§ DECEMBER 1967 cline; in the Pacific the increase is fourfold (to 16 * 10-1* g/liter} and in Atlantic and Pacific Oceans Shipboard analysis of the concentration of radon gas in samples of seawater offers three important types of information: (i) the distribution of Ra2?6 in the world ocean, (ii) the ex- smooth increase in radium takes place downward through the main thermo- is negligible, compared to that in surface seawater, radon continually escapes from the sea. By analyzing the vertical distribution of radon deficiency in the sutface ocean it is possible to determine the rates of both vertical mixing and gas exchange. Water in the pores of deep-sea sediments contains 10+ to 105 times more radon than the overlying seawater. Hence, radon diffuses from the sediment into the sea. The vertical distribution of excess radon in near-bottom water provides an index of. the rate of vertical mixing. This report extends an earlier study (J) by present- ing positive results for each of these applications. Radon was extracted from 20 to 40— liters of seawater by circulating He gas in a closed system at the rate of 2 7 North Atlantic, twofold (to 8 x 10-g/liter). Two explanations have been offered for the deficiency of radium in surface relative to deep water. Koczy (2) suggested that it reflected radioactive decay during the period in which surface water was isolated from the deep sea. The mixing rates required by this hypothesis are an order of magnitude lower than those that explain the vertical distribution of natural radiocarbon [see Broecker (4)]. Chow and Goldberg (5) have suggested that the deficiency is generated in much the same way as that for silicon and phosphorus. Radium in surface water is fixed onto particulate matter that sinks to the deep ocean where the radium redissolves. Their demonstration that, in the Pacific, barium shows a fourfold enrichment in deep relative to surface water strongly supports this alternate hypothesis. Despite the difference in the deep-tosurface anomaly for radium in the two oceans the same particulate extraction rate, /, is required for both. Material balance requires that 7 = R(Cp — Cs). The rate of transfer, R, of water across the main thermocline is given by R= h tu(Cp’ — Cs'/Cop’) where ft is the mean depth of the ocean; Cp,’ and Cs’ are the C'#/C ratios in the deep and surface ocean; 1307