GEOCHEMISTRY AND PHYSICS OF OCEAN CIRCULATION 303 



the fact that it is added to the bottom rather than the top of the 

 ocean as is the case for other tracers. 



Since, to date, only C^^ data are available in sufficient quantity 

 to allow any firm conclusions to be drawn concerning mixing rates, 

 I shall demonstrate the potentialities and limitations of isotope 

 tracers through the consideration of radiocarbon alone. The 

 application of C^^ to oceanographic mixing problems was originally 

 proposed by Ewing and Kulp at Lamont. It has since been 

 successfully used by workers at the Scripps, the New Zealand, and 

 the Stockholm radiocarbon laboratories as well as at Lamont. 



The results I shall give are largely unpublished. They have been 

 accumulated over the past four years at Lamont. A large portion 

 of the Lamont staff has contributed to the shipboard and labora- 

 tory phases of this work. Drs. Ewing, Gerard, Heezen, and 

 Takahashi have directed the difficult and hazardous operation of 

 collecting 400-liter samples of water and Drs. Kulp, Tucek, and 

 Olson have cooperated at various times in developing various 

 phases of the laboratory procedure. 



The procedure used is illustrated in Fig. L The sample is 

 collected in a steel barrel which is continuously flushed as it is 

 lowered into the sea. It can be sealed off at the desired depth by 

 dropping a messenger weight from the surface. The water is 

 processed on the deck of the ship by acidification and bubbling 

 with COo-free air or N2. The CO2 released is absorbed in KOH. 

 The yield is checked on the ship by acidifying a small portion of 

 the KOH and measuring the pressure of the CO2 released with an 

 oil manometer. Upon return to the laboratory, the CO2 is released 

 from the KOH by acidification, and purified by absorption on 

 CaO. The radioactivity of the purified CO2 gas is assayed in a 

 large volume proportional counter with massive iron shielding and 

 anticoincidence shielding. The CO2 gas is also analyzed in a mass 

 spectrometer in order to determine the concentration of the stable 

 trace isotope C^\ 



As shown in Table II, the results are first expressed as per mil 

 difference of the radioactivity and hence C^^ concentration in a 

 sample from that in a standard. In order to eliminate differences 

 resulting from isotope fractionation, all results are normalized to 



