fell than could be contained in the tanks and much waslost through overflow. The results are based on the water remaining in the tanks at the end of the collection period. Floccing procedure. The rainwater in the tanks was stirred with a jet of compressed air and a solution of alumi- num sulfate was added to make the final concentration about 0.2 gm Al.(SO4)3:18 HO per gallon of water. The pH was adjusted to 7.0 + 0.1 with a sodium carbonate solution using LaMotte standards with bromthymol blue indicator. Stirring was continued for 45-60 min and the treated water allowed to settle undisturbed overnight. After settling, the clear supernate was removed by syphoning and the Al(OH); floc transferred to glass bottles for further settling and concentrating. The floc from 1,000 gal of water could generally be concentrated to a volume of 2-5 gal. This procedure has been checked by adding 50-100 mg of soluble lead and bismuth salts to 100 gal of water and then collecting them with aluminum hydroxide floc as described above. At the concentrations used (0.6-1.1 * 10-8 mole/liter}, recoveries were about 70-90%, including all losses due to separation of the floc and chemical separation of the material. Chemical procedure. The floc samples were boiled with hydrochloric acid (1:4) and filtered from anyinsoluble residue. In some cases this residue was further treated with nitric, sulfuric, and hydrofluoric acids and analyzed separately. This treatment was discontinued wheninsignificant quantities of RaD, RaE, and Po activities were found to be associated with this material. Carriers equivalent to about 75 mg each of BizO; and PbSO, were added to the solutions and equilibrated with the radioisotopes. Standard procedures for separating bismuth and lead were used. The separated materials were redissolved and reprecipitated to improve the radiochemical purity.* The time of separation was noted and the samples ignited, weighed, and from the analysis of a few rainwater samples collected at Washington, D.C., are shown in Table 1. The identity of the RaD (22-yr Pb?!") was established by following the rate of buildup of Rak (5-day Bi*!°) in the lead fraction. The RaE contained in the bismuth fraction deeayed with a 5-day half-life. Some a-activity due to Po#!® was invariably found in the bismuth fractions. The separation of Po?! in earrier-free condition gave activities as high as 5,000 dpm from several collections; however, this procedure was not employed routinelysince it gave only qualitative results. A number of studies were made on the distribution of the RaD in different Vol- Location, date collected (1960) Glenview,Ill. Jan. Jun, 2-3 Jun, 24-30 Jul. 6-19 Panama Mar, 7-14 Mar, 16—Apr. 24 May 18-27 Rainfall (in.) 1.6 (squall) Flash showers 1,23 (showers) 0.48 15.96 (continuous) Jun. 28-Jul. 20 10.78 Jul. 20-Aug. 14 16.17 Nov. 24-Dec. 15 20.27 Jan. 5-Feb. 6 (51) 1.75 Hawaii Mar. 8—Apr. 7 May 2-Jun. 12 Aug. 13-Sep. 18 Sep. 18-Dec. 11 Philippine Islands May 4-25 Jul. 4-Aug. 2 Aug. 2-30 Oct. 3-25 Oct. 30-Nov. 30 Samoa Apr. 1-30 May 1-31 Jun, 1-30 Jul. 1-31 Aug. 1-31 (continuous) (continuous) (continuous) (showers) insoluble material and the highest concentration of activity. Collections of RaD activity made at a numberofdifferent sites are compared in Table 3. Some variations are to be noted at each site though they are not as great as the differences between col- lections at different locations. Rain- water collections from inlandsites such as Glenview, Ill., and Washington, D.C., exhibit the highest concentration 265 Insol- uble residue (gm) * Car- rier recovery (%) 89 B-activity of Bi20, fraction} dpm dpm/gal 6,600 25 564 563 38.9 95 76 13,100 413,700 23 24 563 6.6 22 5,640 10 170 0.1 102 500 0.3 87 250 500 500 500 500 0.8 470 dpm/gal (avg) 21 2.8 96 2,270 * 0.5 4.5 * 101 96 79 83 1,920 1,210 1,070 1,130 3.8 2.4 2.1 2.3 3 600 9 1.2 107 100 89 90 1,310 7200 4,890 5,340 3.3 «2.1 12 18 ) shuwers continuous cont. for 2 wk showers showers 270 270 270 270 270 * * * * 3.4 152 108 143 94 94 2,090 2,070 2,280 3,060 3,000 8 8 8 11 11 9 daily 400 96.5 204 2,960 daily daily daily daily 400 400 400 400 verted to decomposition rates by applying appropriate geometry corrections, Oct, 23-Nov. 6 Dec. 5-27 0.4 (scattered) 2.1 (scattered) Kodiak, Alaska ume collected (gal) * 0.2 5,0 3=16.7 2.3 3.3 1.9 7.6 * At times when fission-product activity is present in the sample, further radiochemical purification of these fractions is required. eral, however, the first portion of the rain contains the greatest amount of 400 350 400 300 Jun. 6-19 Jun, 19-Jul. 5 Jul. 7-17 Sep. 11-25 Some of the data resulting concentration of activity in the sample and quantity of dirt, character of the rain, or time between rains. In gen- showers showers showers continuous Sep. 1-30 Results. There TABLE 3—Distribution of RaE in Rainwater at Various Locations mounted on plastic planchets for count- ing. The counting rates obtained by use of standard techniques were con- portions of the rain (Table 2). seems to be no correlation between the daily (scattered) (scattered) (scattered) (continuous) 7.4 0.7 0.4 106 90 92 97 95 1,010 400 500 500 500 4.8 1.7 6.8 1.0 90 107 100 89 2,620 2,300 1,550 3,450 6.5 4.6 3.1 6.9 500 500 5.5 4.0 89 101 6,350 2,590 12.7 5.2 400 * 4.5 * Residues dissolved completely by wet ashings and fusions. Tt Corrected to time of Pb-Bi separation. 330 290 460 170 6=6—0..8 0.7 1.1 (0.4 2.5 2 6.5