Variation of scutellar bristles in Drosophila, I. Genetic leakage.

HE number of bristles on the scutellum of Drosophila is characteristically Tfour. Flies occur infrequently which have more or less than four bristles. PAYNE (1918) selected for increase of the number of scutellar bristles, producing, after several generations of selection, flies with many extra scutellars. This showed that there was considerable unexpressed genetic variation for number of scutellar bristles in flies with four bristles. He further showed that a major component of this variation was sex-linked. scute alleles are characterised by a reduction of the number of scutellar bristles, but the reduction is not constant. scute flies vary in their number of scutellar bristles. RENDEL (1959) selected for increased and decreased number of scutellar bristles in scute flies, producing a low line in which almost all flies had no scutellars, and a high line in which almost all the flies had four bristles, and some had extra scutellars. Substitution of the + for the sc allele in the high line showed that selection for scute modifiers had affected the expression of the + allele. Wildtype flies of the high line had extra scutellars. Both PAYNE’S and RENDEL’S results can be explained by the existence of a canalisation of development at four bristles (see RENDEL 1959). Canalisation of extreme phenotypes have been demonstrated and discussed by WADDINGTON (1956) and MILKMAN (1960). DUN and FRASER (1958) demonstrated that the number of secondary bristles in mice is canalised at an optimum number, using the Tabby gene to demonstrate unexpressed variability of whisker number in + mice. On the assumption that there is a canalisation of bristle development at four bristles, then the infrequent wild-type flies with more or less than four bristles are caused by segregation of the scutellar genes producing combinations with a summed action of more or less than four bristles. This can be termed “genetic leakage” of the canalised scutellar genotype. The term is introduced to distinguish these phenomena from that of “phenodeviants” as discussed by LERNER (1954). In this paper data are presented of genetic leakage in wild and laboratory populations, and of comparisons of scute and + variability. Patterns of bristle position: PAYNE (1918) and RENDEL (1959) in their studies of extra-scutellars make no point of differences of position of extra or missing