instead, the calculated collection efficiency determined by Equation 1 for each sample should be used. Comparing the Pu results in Table 2 for the reef and groundwater, however, we find large discrepancies in the concentration determined by the MICE and radiochemical methods. This becomes even more apparent if we compare the calculated efficiency of the MICE samples and the measured efficiency using the radiochemical data. The other notable difference in Table 1 and Table 2 is the range of activity in the 1 pm prefilters. In the ocean and lagoon samples, the particulate Pu ranges between 8 and 64% (average 26 + 16%) of the total sample concentration. The concentration range of particulate Pu for the reef and groundwater was 12 to 98% of the total sample activity with an average of 67 + 28%. The higher calculated collection efficiency for the reef and groundwater samples could be caused by microparticulates penetrating the prefilter and being sorbed on the first MnO» cartridge. If this were the case, then, the measured collection efficiency (value of A/RC in Table 2) should also be higher than the lagoon samples. But this is not the case. Plutonium associated with microparticulate or colloidal materials, if present in the reef and groundwater, is not Ppreconcentrated significantly by the first MnOo cartridge. In the seven reef and groundwater samples, an average af only 334% of the plutonium measured by radiochemistry was accounted for by the MICE procedure. This may be an indication that the reef and groundwater contain plutonium in chemical forms that are not readily exchangeable with the MnOo cartridge. An attempt was made to further characterize the Pu in the reef water by passing aliquots of filtered samples through the four types of ion exchange resins listed in Table 3. The cation exchange resin, AG-50, retained 28% of the plutonium; the two anion exchange resins, AG-1 and AG-21K, adsorbed 23-28%; and the Chelex100 adsorbed essentially 100% of the total filterable amount determined on a separate aliquot by radiochemistry. We also estimated the plutonium associated with the organic materials in the water by incorporating a charcoal cartridge in line with the MICE samples. In one sample, 4% of the plutonium activity in the filtrate was found in the charcoal when it was placed ahead of the MnO» cartridges and 3% was adsorbed when the charcoal cartridge was placed behind the MnO, filters. This small but distinctly organically bound fraction of Pu was not adsorbed by the MnO> cartridges. The high adsorption shown by the chelating resin, Chelex-100, suggests that the major fraction of plutonium in the filtrate is chemically reactive. The high Chelex-100 value also indicates that inorganic species different from that found in the lagoon, perhaps strongly complexed, are the probable form of plutonium in the reef water. 589