The frequency-dose relation of x-ray-induced y-suppressed lethals in Drosophila.

E A R L Y in radiation genetic investigations it was reported that X-ray-induced sex-linked recessive lethals were not homogeneous in composition. Instead they included: ( 1 ) apparent intragenic changes without observable cytological abnormalities (point mutations) ; (2) deficiencies; and (3) changes associated with and, for the most part, inseparable from chromosome rearrangements (OLIVER 1932; DEMEREC 1937). It was also reported that recessive lethals induced by X rays increased linearly with dose (OLIVER 1932; TIMOF~EFFRESSOVSKY 1939). More recently EDINGTON (1956) reported that X-ray-induced sex-linked recessive lethals increased more rapidly than expected on the basis of linearity, but OSTER (1958) reported no departure from linearity. Although intragenic changes and small deficiencies increase linearly with dose, gross deficiencies and chromosome aberrations have been shown to increase as the 1.5 power of the dose (BAUER 1939; TIMOF~EFF-RESSOVSKY 1939; MULLER 1940). If the aberration-associated lethals were dependent on the chromosome aberration for their expression (position effects) as MULLER advocates (MULLER and PROKOFYEVA 1935; MULLER 1940, 1950, 1954), X-ray-induced recessive lethals should increase faster than expected on the basis of linearity. Since this was not observed, LEA and CATCHESIDE (1945) and HERSKOWITZ (1946) independently postulated that recessive lethals arise as the result of chromosome breakage, with few, if any, originating as point mutations or position effects. This hypothesis was based on the observation that the frequency of X-rayinduced chromosome breaks increased linearly with dose ( SAX 1940; CARLSON 1941 ) . Intragenic changes were considered therefore, to have arisen as the direct result of breakage with subsequent restitution of the broken ends, and aberrationassociated lethals owed their expression to the breaks that by chance became involved in a chromosome rearrangement. Under these conditions the frequency of lethals induced by X rays would be expected to increase linearly with dose in the same manner as chromosome breaks. FANO (1947) and MULLER (1950) pointed out, however, that the recessive lethal data are at variance with results expected on the breakage hypothesis of lethal origin. It was explained that breaks increase linearly with dose, but the incidence with which these breaks recombine in inviable combinations (domi-