Three observations should be made from the analysis of 4 samples of a Yugoslavian population: (1) The sero-anthropological heterogeneity and the ne cessity of conducting multiple investigations to objectively learn the differential distributions of the phenotypes car ried by immunoglobulins should be realized. (2) The existence of a Gm(21) subspecificity should be observed. (3) The necessity of exhaustive familial analysis should be used to disclose the genetic mechanisms which determine the immunoglobulin phenotypes.