606 CYCLES OF ORGANIC AND INORGANIC SUBSTANCES 



similar curves may be fitted between 1600 and 1850 m, but they 

 follow from the thesis and are not evidence for it. 



This paper was presented on 1 October, 1959, at Copen- 

 hagen at the meeting on the IGY preceding the International 

 Council for the Exploration of the Sea. This sandwich-like struc- 

 ture was criticized during discussion on the grounds that if it 

 existed it would have been recognized earlier. In reply one may 

 say that it is only in very recent years that we have had the 

 means by precision titration of chlorinity (Hermann, 1951; Bather 

 and Riley, 1953) or by measurement of conductivity (Schleicher 

 and Bradshaw, 1956) to obtain sufficiently refined data. Oxygen 

 data also ha\'e had to be handled in a somewhat unusual way in 

 order to yield the necessary precision. 



There is no evidence at any depth of an apparent density in- 

 version such as we and others had suspected on the basis of less 

 precise Knudsen titrations. 



It is of interest that extrapolation of the abyssal curve leads 

 to the point: potential temperature, —0.9°; salinity 34.67%; 

 closely corresponding to the properties of the Antarctic bottom 

 water in the Atlantic- Antarctic basin. This provides a simple 

 means of estimating the amount of the Antarctic component in 

 the eastern North Atlantic. Moreover, the nature of the curve 

 suggests that water with an Antarctic component mixes upwards 

 and in the Bay of Biscay vanishes sharply at 2800 m depth. 



Relation between Potential Temperature and Oxygen 



W^hen our results were last reported part of our work had to be 

 withdrawn at the last moment since we had established that in 

 some of our samples there had been very considerable consumption 

 of oxygen in the water bottles while they were being hauled in. 

 We have now coated our bottles with an epoxy resin, Araldite, 

 and seem to have completely eliminated this source of trouble. 



First let us consider the water below 2500 m at station Cavall. 

 Figure 3 epitomizes the four cruises in 1958. The standard devia- 

 tion between duplicates does not exceed 0.04 ml /liter. Short-term 

 variations of 0.1 ml/liter or so appear at all depths. 



In November on board R.R.S. Discovery II it was conxenient 



