142 NIELSEN [CHAP. 7 



standard 14 C ampoules ready for use., and subsequently counts the radioactivity 

 from the filters with the plankton. 



A well-defined amount of 14 CC>2 (as dissolved NaHCC>3 in sealed 1-ml am- 

 poules) is added to the sea- water before an experiment. The content of CO 2 

 (total) in this water must be determined or estimated. If we assume that 

 14 CC>2 is assimilated at the same rate as 12 C(>2, then by determining the content 

 of 14 C in the plankton after the experiment we also determine the total amount 

 of carbon assimilated. It is necessary only to multiply the amount of 14 CC>2 

 found by a factor corresponding to the ratio between CO 2 (total) and 14 C02 

 at the beginning of the experiment. The amount of 14 C assimilated is deter- 

 mined by measuring the /3-radiation from the plankton retained by a collodion 

 filter. 14 C has a half -life of 5500 years. Thus 14 C ampoules will remain virtually 

 unaltered for hundreds of years. 



There is no reason for describing here all of the different details of the 

 technique ; the literature has already been cited above. One detail, however, 

 should be mentioned. It concerns the type of Geiger-Muller tubes to be used 

 for measuring 14 C. A windowless counter is now used by most workers using 

 this special isotope, and also by many of those working with primary produc- 

 tion. The windowless counter has one definite advantage as compared with 

 tubes possessing end-windows. At a definite rate of radiation from 14 C the 

 number of counts is much higher. A mica-window — even a thin one — absorbs 

 a considerable part of the radiation from 14 C. 



Unfortunately the windowless counter has a definite disadvantage which 

 has given trouble to several of the workers using the carbon-14 technique for 

 measuring primary production (see, for example, Doty, 1959, page 43). It is 

 difficult to determine the self-absorption curve necessary for estimating the 

 activity of the solution found in the ampoules if a windowless counter is used. 

 The present author has, therefore, always avoided the use of such counters. 



For the purpose of counting the activity of the solution in an ampoule, 

 carbon dioxide is precipitated as barium carbonate which is then filtered off 

 by a membrane filter. Of primary importance is the extrapolation of the activity 

 thus obtained to a sample thickness of mg/cm 2 on the filter. This is absolutely 

 necessary, because the weak radiation from 14 C is readily absorbed also by the 

 barium carbonate itself. This loss is called self -absorption. 14 C emits ^S-rays 

 with very varied, although always low, energies. If all /3-rays had the same 

 energy, a curve on semi-logarithmic paper showing the percentage of maximum 

 specific activity as a function of sample thickness would be a straight line. In 

 this case an extrapolation to zero thickness is very easy. If a relatively thick 

 mica-window is used, only the most energy-rich radiation from 14 C is able to 

 penetrate and we obtain an almost straight self-absorption curve. In this way 

 the number of counts registered would be considerably reduced, and con- 

 sequently thin mica-windows must be used instead. Fig. 4 presents three self- 

 absorption curves : one obtained with a windowless counter, one with an end- 

 window counter (1.5 mg/cm 2 ) and one with an end-window counter (3.0 mg' 

 cm 2 ). Although the last curve is nearest to the straight line, the present writer 



