104 reliable concentrations. The other problems encountered involve uncertain- ties in the differential fractionation of radionuclides during sample preparation procedure, The third goal was to attempt to provide an estimation (which was free of some of the uncertainties mentioned above) of the total alpha radioactivity in a single sample. The data for this treatment was provided from the analysis of the "standard"sediment for which 22 different sized aliquots were prepared and counted. This provided, as wel] as a means of reducing the sample inhomo- geniety problems, a semiempirical justification, or check,of the calculated range of the alpha particles in the sample. The alpha radioactivity measured in this “standard" Bravo Crater sanple® was found to be 14.4% (21 pCi/g) less than the 145. pCi/g concentration obtained by summing the measured concentrations of 239+240 Pu, 24 An 23854, 234 daughter radionuclides in the sample. U, and 226 Ra and its alpha-emitting In this one sample, as opposed to the others measured, the differences between the summed alpha activities measured by spectrometric techniques and those measured by using the total alpha technique were not believed to arise from bias introduced by imperfectly prepared or unrepresentative aliquots. The large difference in the concentrations obtained by the two methods can thus be due to the presence of alpha-emitting radionuclides other than those measured individually, or to an incomplete or incorrect treatment of the data. If the latter case is the major cause of the difference in the concentrations found, it is most likely that the error arises from either: (1) fractionation of the radionuclides during the sample prepara-~tion process or, (2) to an error in the computed density (thickness) of the sample which completely attenuates the alpha particles emitted. The dearee of radionuclide fractionation which occurred during the sampte preparation a. L.R.E. No. 25653, No. 5, Ratch 2 of 4, 155-160 ft.