One-half the sample is analyzed for strontium-90 and strontium-89 strontium carrier is added and the strontium is precipitated from first as the carbonate then as the nitrate. Further purification barium chromate precipitation and hydroxide scavenging. After an as follows: the samples is made by ingrowth period, yttrium carrier is added and the yttrium is extracted into TTA (2-thenolytrifluoroacetate) at pH = 5.0 The yttrium is stripped from the TTA with dilute nitric acid, precipitated as the oxalate and beta counted for yttrium-90. The strontium is precipitated as the carbonate and beta counted for total radio-strontium. The strontium-89 activity is calculated from the total radio-strontium measurement after correction for the strontium-90 content. Analysis of the other half of the sample for the plutonium isotopes is briefly ‘described below: Plutonium-236 tracer is added, the plutonium reduced to the +3 state and co-precipitated with lanthanum fluoride. The lanthanum fluoride is converted to lanthanum hydroxide, dissolved in 7.2 M nitric acid and the plutonium oxidized to the +4 state. The solution is passed over an anion exchange resin in the nitrate form. The resin is washed with additional nitric acid and then with 9 M hydrochloric acid. The plutonium is then eluted from the resin with a mixture of 0.36 M hydrochloric acid and 0.01 M hydrofluoric acid, electroplated onto a stainless steel planchet from a sulfuric acid-ammonium = Daas cogARIEP ae chasse ae tn chtae mae Fy ae RE OREN permet . sulfate electrolyte, and counted with a silicon surface barrier detector linked to a multichannel analyzer. With the sample electroplated onto a 3.1 em? area, the counting efficiency is ~314 and the resolution 75 KeV. The minimum detectable quantities are 0.10 fCi/m3, 0.40 £Ci/m3, and 9.003 £Ci/m? for strontium-90, strontium-89 and the plutonium isotopes respectively. The maximum 20 counting error for the strontium-90 is 25 percent of the reported value, although in the majority of cases the 20 counting error is approximately 10 percent of the reported value. The corresponding values for strontium-89 are about twice as high. The maximum 20 counting errors are 50 percent and 67 percent of the reported values for plutonium-239 and plutonium-238 respectively, but are appreciably lower as the results increase above the minimum detectable levels. RESULTS AND DISCUSSIONS The results of strontium-90 analyses of ground level airborne particulates from monthly composite samples are presented tn Figure 1. Difficulties in sampling and analysis caused the loss of two samples (October 1963 and August 1967), and during a change over in personnel responsible for collection and analysis of the data, several months of ambiguous data (September, October, November, and December 1967) occurred. Figure 1 illustrates the expected spring maximum which occurs each year and which appears to be extremely reproducible in terms of time of appearance. A line (solid) indicating the expected spring maximum highs, based on the strontium~90 concentrations in the spring of 1964 and a stratospheric residence of half-time of ten months, is shown. The expected strontium-90 curve (dashed), having the shape of the 1964 curve because no fresh tntrusion of material occurred until the end of that year, is also shown. The difference in the expected and actual curves indicates strontium-90 of post 1963 origin was present.