Widespread resistance to pyrethroids threatens malaria control in Africa. Consequently, several countries switched to carbamates and organophophates insecticides for indoor residual spraying. However, a mutation in the ace-1 gene conferring resistance to these compounds (ace-1R allele), is already present. Furthermore, a duplicated allele (ace-1D) recently appeared; characterizing its selective advantage is mandatory to evaluate the threat. Our data revealed that a unique duplication event, pairing a susceptible and a resistant copy of the ace-1 gene spread through West Africa. Further investigations revealed that, while ace-1D confers less resistance than ace-1R, the high fitness cost associated with ace-1R is almost completely suppressed by the duplication for all traits studied. ace-1 duplication thus represents a permanent heterozygote phenotype, selected, and thus spreading, due to the mosaic nature of mosquito control. It provides malaria mosquito with a new evolutionary path that could hamper resistance management.

number of X-autosome translocations have been described in the mouse, A which in the heterozygous female show a variegated phenotype for autosomal genes attached to the X chromosome ( CATTANACH 1961 ;RUSSELL and BANGHAM 1961 ; RUSSELL, BANGHAM and SAYLORS 1962). The primary cause in each instance was considered to be the inactivation of the associated autosomal genes (CATTA-NACH 1963;RUSSELL 1964) when the rearranged X is inactivated in the course of the normal process of X-inactivation (LYON 1961 ) . Evidence substantiating this hypothesis for one of the translocations has been obtained using both genetical and cytological tests (LYON 1963;OHNO and CATTANACH 1962).Several of the rearrangements have involved the translocation to the X of autosomal regions carrying more than one marker gene and here it has been shown that there is a secondary mechanism responsible for the variegation. Autosomal loci remote from the break point tend to be less influenced by the inactivation process than those lying closer to the break point (CATTANACH 1961;RUSSELL 1963), a phenomenon which appears to be analogous to the "spreading effect" observed in Drosophila V-type position effects (see review by LEWIS 1950) . Two clear mechanisms responsible for mouse translocation-induced variegation thus exist, (1 ) that due to the random inactivation of the rearranged and normal X in each cell, and (2) that caused by the Drosophila-type position effect which operates only when the rearranged X is in its inactive, heterochromatic condition. Variegation due to the second mechanism alone may be observed in situations where the randomness of X-inactivation is suppressed ( CATTANACH 1966a).We have recently been able to show that the position effect variegation is under genetic control (CATTANACH and ISAACSON 1965) ; genotypic selection for large and small amounts of albino areas in the coats of mice showing a translocationinduced variegation for albino was successful in establishing two lines of mice, one in which the autosomal (albino) locus was inactivated in every cell in which the rearranged X was inactivated. and another in which this was not always the case, i.e., the proportion of albino hairs was 50% in one line and 30% in the other. The data indicated that only one or a very few factors were responsible, and further investigations revealed that they were located in the rearranged X chromosome itself ( CATTANACH 1966b).