4 2. ‘Che rare variants. --In Table 2 we have summarized not anly our eum findines but ulso the results of all the other studies of xare vactants of 24 systems iu the Literature. to locate Any effort to treat "rare'-variants involves same aicbitrary decistons; at this time. in Micronesians which we huve been able no approach is apt to find universal acceptance We exclude from this summary any variant which for the totality of the representatives of the population studied to date, occurs in more than 2.0% of the group (one of the conventional definitions of a polymorphism). By this definition of rare variant, we exclude from the tabulation the polymorphisms involving types 3 and 7 of the PCM system reported by Blake et al (12) in the Western Caroline Islands, and the polymorphism for the type 2 of phosphoglycerate kinase reported by these same authors. This same convention willrequire us to eliminate fiom the summary of rare variants in Amerindians (see below) the Yanomama-2 variant of albumin, a variant thus far, encountered in a single tribe, but there with a gene frequency of 0.08. In sampling populations where rather close biological relationships between individuals can scarcely be avoided, it is well to remember that even in samples of -1000, a variant limited tu inembers of a single extended kindred may asyume the proportions of a polymorphism as here defined. In the case of the variants excluded from the Micronesia count, a variant with the electrophoretic mobility of Lype 7 PGM, 1 has bean encountered at several localities in the Far East, avd it seems likely this is a "widespread" polymorphism. However, the same cannot be said for the variant with the mobility of PCM, -3 or the variant with the mobility of PGK-2. morphisms. They may be “private'" poly- Lf£.further gtudles in Micronesia revealed these. alleles