CHAP. 34] THE PLEISTOCENE RECORD 917 



Pacific using radium as an index of 23oxh. Although radium was later shown to 

 migrate rather freely through deeji-sea sediments (Pettersson, 1951; Kroll, 

 1954, 1955), radium migration is relatively unimportant in sediments with a gross 

 rate of sedimentation greater than about 1 cm/ 1000 years (Koczy and Bourret, 

 1958), which is generally the case for Globigerina ooze. If the stratigraphy of the 

 Atlantic and Caribbean cores dated by Urry (Bramlette and Bradley, 1940; 

 Cushman and Henbest, 1940; Cushman, 1941) is interpreted in the light of 

 present knowledge, it appears that the last major warm stage was dated from 

 about 110,000 to about (55,000 years B.P., the last major cold stage from 

 65,000 to 12,000 years ago, and the postglacial from about 12,000 years ago. 

 These figures have been substantiated by recent dating by i^Q and ^sipa/sso^h 

 methods (cf. Rosholt et al., 1961, 1962). It is unfortunate that stages older than 

 the last major warm stage cannot be clearly recognized in the Cayman core 

 dated by Urry (Piggott and Urry, 1942; Cushman, 1941), a result, probably, of 

 much post-depositional solution of the foraminiferal shells having occurred at 

 the depth from which the core was raised (4800 m). 



The difficulty associated with possible changes in rate of supply of uranium to 

 the sea and changes in the rate of clay sedimentation is eliminated by using the 

 ^3iPa/230Th method, because such changes would affect equally both parent 

 uranium isotopes. Using this method, Rosholt et al. (1961, 1962) dated temj)era- 

 ture fluctuations back to about 175,000 years ago in three apparently undis- 

 turbed Caribbean cores. The ages obtained from the younger jDortions of these 

 cores correlate with the ^^C chronology within the limits of error of the two 

 methods, ^sipa/saoxh dating of a North Atlantic core (Swedish core 280) gave 

 ages that are about 30,000 years greater than the ages obtained from the 

 corresponding levels of the Caribbean cores, and do not correlate with the ^^C 

 chronology in the region of common range. This discrepancy is believed to result 

 from the introduction of reworked clay into the Atlantic core (Rosholt et al., 

 1961). 



If the ages of the temperature stages of the deep-sea cores obtained with the 

 various methods mentioned above are compared with the ^^C chronology of the 

 continental stratigraphy (Flint, 1957, p. 395; Andersen etal., 1960), it appears 

 that stage 2 (see Fig, 6) represents the Main Wiirm; stage 3 the Early Wiirm- 

 Main Wiirm interval; stage 4 the Early Wiirm; and stage 5 the last interglacial 

 (cf. Emiliani, 1955, 1958; van der Hammen, 1957). Thus the correlation be- 

 tween core stages 1 to 5 and continental stages from the present to the preceding 

 major interglacial seems to be good. The tentative correlation between older 

 core stages and continental stages suggested by Emiliani (1955) is still con- 

 sidered speculative, as it was when first proposed. 



Oxygen-isotopic analysis of calcareous benthonic Foraminifera from the 

 section of core 234 representing the last interglacial showed that bottom tem- 

 perature followed closely the pattern of surface temperature, but with con- 

 siderably smaller amplitude (Emiliani, 1958). These and other analyses of 

 calcareous benthonic Foraminifera (Emihani, 1955, 1958) showed that the 

 maximum glacial-interglacial temj)erature range of the bottom water was on 



