136 HISTORY OF THE OCEANS 



the sediment, and a lower average carbonate content, indicating a 

 general decrease in the productivity. 



It has been demonstrated by Emiliani that the strata formed at 

 increased productivity in the equatorial Pacific are associated with 

 colder surface water, as indicated by the oxygen isotope ratio in 

 planktonic Foraminifera. He has also shown that during each 

 given time period the surface temperature of the ocean drops 

 toward the equator, consistent with the upwelling of cold deep 

 water there, and that this temperature gradient becomes steeper 

 during periods of high productivity (Fig. 2, top graph). 



All these observations strongly suggest that the periods with a 

 high rate of accumulation of organic remains were periods of 

 increased rate of vertical water transport in the equatorial diver- 

 gence, which ultimately appear to be caused by increase in the 

 trade wind intensity. The ocean sediments thus provide us with a 

 record not only of polar wandering but also of the intensity changes 

 in the low-latitude atmospheric circulation, which is directly tied 

 in with the great climatic changes that during the last geological 

 period resulted in the ice ages. 



The direct effect of cooling of the surface water at intermediate 

 latitudes has also been traced over large areas in the Atlantic 

 Ocean and in the Mediterranean through the original work by 

 Schott (1935) on species of Foraminifera, sensitive to cold and 

 warm water, followed by a number of similar studies by Cushman 

 and Henbest, Phleger, Ovey, Ericson, and others. Quantitative 

 information on the temperature development has been provided 

 by Emiliani's (1955) extensive studies of the paleotemperature as 

 indicated by the oxygen isotope ratio in tests of planktonic 

 Foraminifera. Again it is here evident that the last period of 

 cooling coincides in time with the last ice age, and by inductive 

 reasoning it seems safe to assume that the previous low-tempera- 

 ture periods correspond to the older ice ages. The reason why this 

 is difficult to prove directly is that the time scale of these events, 

 as they are recorded on the continents, is only approximately 

 known. 



It should be mentioned in this connection that Emiliani (1955) 

 has attempted to correlate the succession of cold and warm periods 



