population stuly as possible) and an understauding of the structure of the populations oa which the estimates are to be based. This is, oF coucse, crue for plant and animal as well as human populations. fn tun present instanea, we ave of course intrigued by the possibility ‘Chat the iia of these Factores differs in Micronesians from the other ot groups cited, Okey lower mutation rates or greater stochastic loss of new rmatanes in the Tlicronesians. Before these possibilities can be profitably pursued by the indirect approach, sample size must be adequate. There is clearly some’ minimal sample necessary to a trust- worthy estimate of the frequency of rare variants in populaClons like Micronesians and Anierindians, and most investigators would probably auree that the Micronesians do not yet approach that minimum. Un- fortunately, our kaowledge of the clustering of specific rare variants in celutively undisturbed populations is still so scanty that further expettoence is necessary prior to setting that reasonable minimum. SUMMARY Blood specimens from a sample of 187 Marshall Islanders were studied with reference to variants of 22 serum proteins and erythrocyte enzymes. Six of the trafts studied exhibited genetic poly- morphisms (adenosine deaminase, phosphoglucomutase,, acid phospha-~ 1 tase, 6-phosphogluconate dehydrogenase, haptoglobin, group specific component). There was in addition one "rare" variant (of albumin) in 4,047 determinations, These results on rare variants have been combined with those of others on Micronesians and the frequency of rare variants in Micronesiany compared with the frequencies in West I0049b3b