As sea water apart from liquid ammonia has the highest heat capacity of any solid or liquid the deposits collecting on the deep-sea floor, in favourable localities, give a far better record than the land of past temperature changes, provided the dominant component is planktonic Foraminifera, and that the rate of sedimentation of at least one of the other components has remained constant with time. There are three possible methods whereby past temperature changes in the upper layer of the Equatorial Atlantic Ocean can be revealed. The first is the productivity method. As the minimal factor which influences the productivity of planktonic Foraminifera is apparently temperature, it is possible in the ideal case where the rate of sedimentation of the non-calcareous components has remained constant with time; and where there has been no contribution to the carbonate content other than by the shells of planktonic Foraminifera, and provided there has been no appreciable solution of carbonate, to follow the changing temperature in the upper layer of the sea by determining the CO
2
content in a series of samples throughout the length of the core. This method is clearly applicable to the more general case where the rate of sedimentation of only one of the non-calcareous components has remained constant with time. A new technique has been developed for determining accurately the CO
2
content in globigerina ooze cores. The second method, due to Mr Ovey, depends on the species distribution of planktonic Foraminifera in 1000 specimens > 127
μ
. The third method has been developed by Professor Urey and depends on the
18
O content of individual planktonic species. Consideration is given to the possibility of the CO
2
changes being spurious and unrelated to temperature changes. Perhaps the most convincing argument against this hypothesis is the similarity between the carbonate curve in the Atlantic core and the carbonate accumulation curves for the Pacific cores, as well as in the number of maxima and minima and in their respective ages. In the top portion of an undisturbed pilot core, there are apparently CO
2
oscillations of a shorter period. A continuous series of sections, approximately ½ cm thick, were taken down this core. It has been possible to determine the total weight of non-calcareous components, dried at 105° C, in a column of unit area down to any depth in the core, and by correlating two distinct CO
2
oscillations with two climatic changes of known ages, the mass contribution per year can be computed. The inverse relation between TiO
2
, Fe
2
O
3
and CaCO
3
suggests that there has been no marked deviations in the rate of sedimentation of the non-calcareous components with time. On this assumption, it is possible to compute the age at any depth. There is an apparent agreement between the ages of these oscillations and the ages of known second-order climatic changes. The fact that the age according to these computations of the top of the core is A. D. 1838 gives support to these correlations. There are two effects to be clearly distinguished: first, a long period change and secondly, minor oscillations superimposed on these major changes. The necessity for the development of a simple coring device to take short but wide undisturbed cores, which are truely representative of the natural sedimentary column of the deep-sea-floor, is pointed out.
Control of organic carbon accumulation in the Late Quaternary equatorial Atlantic (Ocean Drilling Program Sites 664 and 663): Productivity versus terrigenous supply