Ohno has shown that in somatic cells of human females, the two X-chromosomes are not alike. One behaves in exactly the same manner as the autosomes, remaining in an extended state during interphase and prophase, while the other assumes a heavily condensed state, forming the Barr sex chromatin body. In male somatic cells, on the other hand, the single X never manifests positive heteropycnosis; it always appears fine and elongated. It seemed to us entirely possible that the female X which condenses during interphase is genetically inactive. The active chromosome could not always be derived either from the mother or from the father, since this would give a pattern of inheritance of sex-linked mutations quite different from the pattern as we know it to exist. It seems necessary, therefore, to assume that there is, at least for some period of time during development, randomization of the active and inactive chromosome among the dividing cells of the body. Women would then be composed of a mosaic of cells, some with a functional paternal X-chromosome, others with a functional maternal X-chromosome. Glucose-6-phosphate dehydrogenase (g-6-pd) deficiency, an inborn error of metabolism of humans,9 represents an ideal test system in which to examine this hypothesis. Many heterozygous females with this disorder have red cell g-6-pd levels which are approximately 50 per cent of normal. If the precursors of a heterozygous female's erythrocytes were, as our hypothesis suggests, a mosaic consisting of some cells with the maternial X-chromosome active, others with the paternal X