908 EMILIANI AND FLINT [CHAP. 34 



carbonate deposition, the change being inversely related to temperature. If the 

 oxygen-isotopic composition of the water is known, the temperature existing 

 during the time of carbonate deposition can be determined within one degree 

 centigrade. If the composition is not known but can be reasonably estimated, 

 as in the case of open oceanic water of the not too distant past, reasonably 

 reliable temperature data can still be obtained. 



Paleotemperature analysis has been rather extensively applied to the study 

 of Cenozoic deep-sea cores (Emiliani, 1954a, 1955, 1955a, 1956, 1958; Rosholt 

 ef al., 1961, 1962). Most of the isotopic work has been done on shells of pelagic 

 Foraminifera, which provide data on the oceanic surface temperature and on 

 its variation through the time represented by the deep-sea cores. No post- 

 depositional recrystallization or ionic diffusion through the solid state of the 

 foraminiferal shell material has been noticed. Since empty shells of pelagic 

 Foraminifera are buried on the ocean floor where temj^erature is very low, 

 recrystallization and ionic diffusion would tend to increase the i^o/ieQ ratio 

 in the CaCOs. The result would be temperature readings lower than the original 

 temperatures or equal to the bottom temperature for total exchange. It was 

 found, instead, that shells of pelagic Foraminifera of a Middle Oligocene age 

 from a deep-sea core from the subtropical Atlantic Ocean gave temperatures 

 around 29°C, although the shells had been maintained at much lower tempera- 

 tures for about 30 million years (Emiliani, 1956). Similar results were obtained 

 from two Miocene cores (Emiliani, 1956). The aj)parent preservation of the tem- 

 perature record may be due in part to the fact that diffusion through the solid 

 state is not only very slow but is also strongly dependent on temperature (see 

 Urey et al., 1951, for data on calcite). Thus the low temperature at which the 

 foraminiferal shells were maintained may have been important in preserving the 

 temperature record even though the shells are thin and j^erforated by canals. 

 Organic coatings may also be important. 



Because different species of pelagic Foraminifera appear to deposit their 

 CaCOs at different depths (Emihani, 1954, 1955, 1958; see also Bandy, 1956; 

 Bandy and Arnal, 1957; Polski, 1959) and/or at different seasons (Emiliani, 

 1958; Be, 1960; Phleger, 1960), and because temperature varies rapidly with 

 depth near the ocean surface, and also with time of the year in regions outside 

 the equatorial belt, oxygen-isotopic analysis of deep-sea cores must be per- 

 formed on monospecific samples and different temperature graphs must be 

 constructed for each different species. The superficial species Globigerinoides 

 rubra and Glohigerinoides sacculifera yield a temperature record of greater 

 amplitude and detail because climatic variations affect more surface water than 

 the water below. 



The carbonate component of Globigerina ooze may be conveniently divided 

 into two size fractions, one larger than 62 ]x and the other smaller. In general, 

 the fraction larger than 62 \x consists almost entirely of shells of pelagic Fora- 

 minifera, while the smaller fraction consists of minute shells and shell fragments 

 of pelagic Foraminifera, and especially of coccoliths. In cores of Globigerina 

 ooze from the equatorial, tropical and subtropical Atlantic Ocean, and adjacent 



