[Illustration: _Studies at the California Institute of Technology furnish information on the nature of radiation effects on genes. The experiments produced fruit flies with three or four wings and double or partially doubled thoraxes by causing gene mutation through X-irradiation and chromosome rearrangements. A is a normal male_ Drosophila; _B is a four-winged male with a double thorax; and C and D are three-winged flies with partial double thoraxes._] [Illustration: Four-winged male with a double thorax] [Illustration: Three-winged fly with partial double thoraxes] [Illustration: Three-winged fly with partial double thoraxes] Mutations can be brought about in the sex cells, too, of course, and when this happens it is succeeding generations that are affected and not merely the exposed individual. Indeed, where the sex cells are concerned, the relatively mild effect of mutation is more serious than the drastic one of nondivision. A fertilized ovum that cannot divide eventually dies and does no harm; one that can divide but is altered, may give rise to an individual with one of the usual kinds of major or minor physical defects. The effect of high-energy radiation on the genetic mechanism was first demonstrated experimentally in 1927 by Muller. Using _Drosophila_ he showed that after large doses of X rays, flies experienced many more lethal mutations per chromosome than did similar flies not exposed to radiation. The drastic differences he observed proved the connection between radiation and mutation at once. Later experiments, by Muller and by others, showed that the number of mutations was directly proportional to the quantity of radiation absorbed. Doubling the quantity of radiation absorbed doubled the number of mutations, tripling the one tripled the other, and so on. This means that if the number of mutations is plotted against the amount of radiation absorbed, a straight line can be drawn. It is generally believed that the straight line continues all the way down without deviation to very low radiation absorptions. This means there is no “threshold” for the mutational effect of radiation. No matter how small a dosage of radiation the gonads receive, this will be reflected in a proportionately increased likelihood of mutated sex cells with effects that will show up in succeeding generations. In this respect, the genetic effect of radiation is quite different from the somatic effect. A small dose of radiation may affect growing tissues and prevent a small proportion of the cells of those tissues from dividing. The remaining, unaffected cells take up the slack, however, and if the proportion of affected cells is small enough, symptoms are not visible and never become visible. There is thus a threshold effect: The radiation absorbed must be more than a certain amount before any somatic symptoms are manifest.