roughness effects, and neglect of very low energy emitters. As discussed in references 2 and 3, the field spectrometric exposure rate estimates are very insensitive to local inhomogeneities in isotope concentration, small errors in Sete at relaxation length, or slight ground roughness effects as opposed to the large errors which would be obtained by calculating the exposure rates using concentration data from an atypical soil sample. This is due primarily to the fact that the spectrometer "sees" a large area of soil and averages out most of these inhomogeneities. The estimates of the percentage of the total exposure rate due to each emitter obtained from both the soil concentration data and the field spectrometric data agree quite well and these estimates, whenever given, are probably fairly reliable keeping in mind that on many of the islands they are based on just one soil sample from a single location. Error in TLD Results - The accuracy of the exposures obtained from the TLD data (see Table 1) is best indicated by comparison with the ionization chamber results at mutual sites. The HASL TLD data seem to agree fairly well on the average, although a few individual values appear to be quite far off. The NRDL TLD data appear to be about 20% higher on the average with larger variations. Overall Consistency of Data - The overall consistency of the ionization chamber and spectrometric measurements, TLD Seen eee results, and calculations from the soil analysis indicates that the range of exposure rates on each island and the major contributors to these exposure rates have been determined quite accurately. This consistency is verified by the data in Table 1 and the data discussed in the next section. ats