Determination of total carbon and its radioactivity.

Radioactive carbon has been routinely determined in this laboratory by gas phase proportional counting (1). In this procedure the total COZ is first measured in the Van Slyke-Neil1 manometric apparatus, then transferred to the counting tube where methane is added until the COz-methane mixture reaches atmospheric pressure. The counting is done with a potential of 3800 volts across the counter. Approximately 98 per cent of the disintegrations occurring within the volume enclosed by the silvered area of the counter are measured (l-3). The presence of more t,han 16 per cent of CO2 in the methane-CO* mixture depresses the count, but with the use of a correction curve (Fig. 10 (1)) mixtures containing up to 30 per cent CO2 may be counted with little sacrifice of accuracy. In the present paper, a modification is presented which permits counting at approximately 2000 volts. The U402 is mixed with a gas consisting of 90 per cent argon and 10 per cent methane, commonly called “P-lO.“l The use of the lower voltage has some advantages. Operation at 2000 volts is less likely to give trouble than operation at 3800 volts, particularly in moist weather in a room lacking humidity control, and is therefore likely to require less frequent servicing. Furthermore, well regulated power supplies delivering 2500 volts are more generally available than those delivering 4000 to 5000 volts. The only disadvantage encountered in the use of the P-10 gas in place of methane is that the maximal size of carbon sample that can be used with the P-10 is about 6 mg. of carbon, compared with 16 mg. with the methane. This difference is seldom important, since samples of over 4 mg. of carbon are required only in rare cases, when the specific Cl4 activity is unusually low (see Fig. 12 (1)).