(Reprinted from Nature, Vol. 189, No. 4762, pp. 348-352, February 4, 1961) NATURALLY OCCURRING ALPHA-ACTIVITY OF DRINKING WATERS By R. C. TURNER, J. M. RADLEY and Pror. W. V. MAYNEORD Physics Department, Institute of Cancer Research : Royal Cancer Hospital, London, S.W.3 HE present article represents a continuation of investigations into the nature and levels of naturally occurring x-activitv in man and his environment. We have demonstrated the grat variations of %-activitvy to be found as between different human foods'. Following the observations on foods it was natural to study the radioactivity of drinking waters supplied to the population of Great Britain. Several investigators?-* have reported measurements of the radium-226 content of potable waters in other count.ies, but very few data have been puhlished during recent years on the activity of waters in Britain. Following the observations of J. J. Thomson?early in this century on Cambridge water, there was considerable interest in this subject, but much of the earlier work leaves doubt as to whether the observed values applied to radium-226 or also included gaseous products such as radon-222, known to exist in certain waters at considerably higher levels of activity than its parent. In the present experiments we have measured the radiuin-226 content of 71 drinking waters available in Britain, and in addition we report the values for radon-222 and radium-224 (thorium .Y) when present. In most instances the water samples have been made available to us by the co-operation of the local water undertakings at the request of the Ministry of Housing and Local Government Experimental Procedure Tre water samples were selected from several points of view, including the size of population supplied and the nature of the associated geological formations. The specimens were chosen with the view of assessing the total daily intake vf natural radioactivity throughout Britain from its water supplies. Preliminary experiments indicated that. if polythene containers were used for the colleetion and transport of the samples, it was unnecessary to add either carrier or acid agent to the water at the time of sampling. No residual «-activity could be detected on the walls of such containers even after they had housed specimens of relatively high activity for several days. A litre of each specimen of water to be measured was evaporated to dryness in a ‘Vitreosil’ beaker and the «-activity of the residue measured, using the «-ray counting technique previously described’. In a few eases where the mineral content of the water was very low, 20 mgm. of barium chloride having known low «-activity was added to the water before evaporation in order to provide adequate residue. Measurements were made on each specimen of residue immediatelyafter evaporation andat iutervals of a few days during the following 30 days, by which time the radon-222 and its a-emitting daughters polonium-218 (radium A) and polonium-214 (radium C’) would have reached equilibrium with the radium226 present. After 30 days no further change of a-activity of the residue could be detected in any of the specimens. Assessment of the amount of radon222 above the level required for equilibrium with the ri lium-226 present in each water was effected by observations on a liquid sample 2-3 mm. deep, sealed in contact with a zinc-cadmium sulphide phosphorof area 93 cm.*. Adsorption byzinc sulphide crystals of the daughter products of radon-222 present in solution iad been previously reported by Rosholt", who used the method to obtain quantitative yields of lead-218 (radium B) and polonium-214 (radium C’). It has been observed by us that, when water con- taining dissolved radon at a level of activity many times higher than its content of radium-226 is assembled in such an arrangement, the «-counting-rate increases for the first hour after sealing the sample and thereafter decreases with the half-life of radon-222 (3-8 days), until the activity has fallen to the level represented by the content of radium-226 and other long-lived activitv. If at any intermediate stage the water is removed and the activity of the phosphor itself observed, the counting-rate decreases rapidly during the first 10 min. or sy, corresponding to decay of polonium-218 (radiuin ‘), and thereafter falls with a half-life of approximately 40 min. compatible with the decay of lead-214 (radium 8) and bismuth-214 (radium C) It is thought that the solid decay products of the dissolved radon are deposited on the surface of the zinc sulphide crystals and the «-particles emitted by polonium-218 (radium 4) and polonium214 (radium C’) are assumed counted with 27 afficiency. So long as the water itself is present, the observed activity is maintained with the half-life characteristic of radon, although the radon «-particles themselves may contributelittle to the counting-rate since their range in the medium is only 1 or 2 per cent of the depth of the water layer. Fig. 1 shows the charactcristic pattern of decay of x-activity observed in a water sample containing radon greatly in excess of the amount required for equilibrium with its radium-226 content. Each estimate of radon content has been corrected for decay between the time of collection of the sample and the actual measurement. In a numberof the more active waterresidues, fast pairs of a-particles were observed, indicating the presence of at least the shorter-lived members of the thorium series!®. This conclusion was borne out by the subsequent variation of the thorium series activity with time, since it decreased rapidly during the first 20 days or so following evaporation, after which the thorium activity remained constant at a relatively low level. The presence of radium-224 (thorium -Y) in American potable waters has already been reported by Stehney?, who found the levels of radium-224 in freshly sampled waters comparable with those observed for radium-226. Due to the short half-life of radium-224 (3-64 days) andof its «-emitting daughter members of the thorium series, this x-activity decreases to very low values in the 2-3 weeks subsequent to sampling. Thereis little doubt that the decreasing thorium series activity observed