ON THE DETERMINATION OF THE ACTIVITY IN “C-AMPOULES FOR MEASURING PRIMARY PRODUCTION

A biological method for measuring the activity in 14C-ampoulcs is dcscribcd. All 14COa is assimilated by ChZoreEla and the activity of the algae is measured. The new technique gives results substantially (31%) lower than the values obtained by the original BaCOs technique but seems to be in accordance with other techniques recently developed. When the new technique for measuring the activity of the ampoules is used, the 14C method and the oxygen method give identical results. Skeletonema costatum was used as the test organism.--INTRODUCI’ION In the determination of organic produc-tion by plankton algae by means of the 14C technique first described by Steemann Nielsen ( 1952), it is necessary to measure the radioactivity both of the 14COz added and of the algae at the end of the experi-ment. The activity of the algae is measured after filtration through a membrane filter. The amount of algae is kept so low that self-absorption is negligible. According to the original description of the technique, the CO2 in the ampoules is precipitated as BaC03 and the radioactivity is measured after filtration through a mem-brane filter. Unfortunately, it is generally impossible to make the filtration in such a way that self-absorption is negligible. It was therefore necessary to determine the radioactivity of the BaCOs at zero weight by means of a self-absorption correction curve. The technique for precipitating CO2 as BaC03 and the preparation of a self-absorp-tion correction curve were described by Calvin et al. (1949, p. 104-117). It has always been assumed that the determination of the radioactivity by means of a BaC03-self-absorption correction curve is only an approximation. Two factors work-ing in opposite directions influence the accuracy of the determinations: 1) the wide energy spectrum of the P-rays emitted by 14C, and 2) the self-scattering of the p-radiation. By using a thin-window chamber instead of a windowless chamber, all,&rays of very low energy are cut off, thus reducing the influence of the wide energy spectrum con-siderably. Theoretically, therefore, it should be easier to extrapolate to zero weight, since the curve more closely approaches a straight line on a semilogarithmic plot