Examination, in the salivary gland nuclei of D. melanogaster males, of four cases of translocation between the X and fourth chromosome, involving breaks of the X in widely different positions, disclosed no influence of the rearrangement on the width, morphology or chromatin-staining of either of the separated parts of the X, or on the fourth chromosome. Both parts of the single X were distinctly narrower than were the double major autosomes, as is true for the X of normal males but not for the double X of females. At the junction between a portion of a translocated single X and the double fourth chromosomes the transition in width, morphology and staining was abrupt and striking. As in structurally normal males, however, the parts of a single X, here removed from one another by the translocation, did appear to be somewhat swollen, as compared with half of a double chromosome, but to be correspondingly paler in stain, so as to indicate an unaltered amount of chromatin (see also Offermann, 1936; Rudkin, 1964; Pavan & Frota-Pessoa, 1964).
The above evidence of the regional autonomy in characteristics of the different parts of the X studied by us, and also of the fourth chromosome, is in contrast to the lack of such autonomy found in translocation studies on mammals, where the X chromosome and that joined with it are subject to an influence diffusing along them and thus acting ‘wholesale’, rather than ‘piecemeal’. Likewise, a re-examination of the earlier genetic evidence on dosage compensation in Drosophila leads back again to a decidedly ‘piecemeal’ interpretation of its operation and evolution, according to which most genes in the X, and sometimes even different phases of the action of the same gene, have their own system of separately evolved, scattered compensators, which are also located in the X.
The fact that two so differently working compensation mechanisms as those in Drosophila and mammals have evolved independently to serve the same function emphasizes the importance of that function. That is, it points up the survival value of having the effectiveness of normal genes regulated to a very exact level. For the compensation enables the single representative of the X in the male cell to become equivalent to the two representatives of the X in the female cell. Moreover, this equivalence is of a considerably finer grade than that already afforded by the phenomenon termed ‘dominance’, which has evolved to meet the same basic need (that of phenotypic stabilization), and which has, incidentally, made even the uncompensated effects of one and two doses of either sex-linked or autosomal normal genes not readily distinguishable in most cases.
Taken by itself, the ‘piecemeal’ mechanism of Drosophila provides far stronger evidence for this conclusion than does the ‘wholesale’ mechanism in mammals. For the former must have required the establishment of far more numerous mutational steps and, taken individually, each of these steps was of correspondingly lesser survival value. Since they nevertheless affected fitness enough to become established, it also follows that usually a normal gene—or at any rate one of the kinds whose mutants have usually been studied—confers a significantly higher fitness when not heterozygous for such a mutation in it, despite the seeming recessiveness of most mutations. Thus, the expression ‘normal gene’ continues to have a very high validity.