484 RADIATION BIOLOGY 



complicated combinations include a higher proportion of inviable aneu- 

 centric rearrangements than do the two-break combinations that prevail 

 at lower doses. Since these combinations become lost to view by the 

 genetic methods used, in which only viable offspring are scored, the 

 count of structural changes registered at higher doses becomes still more 

 reduced, in comparison with the number of breaks that actually took 

 part in structural changes at these doses. 



If this is a correct interpretation of the approximate ^^ rule observed 

 in this material, it is not to be regarded as representing a fixed relation. 

 For a lower range of doses than that ordinarily used, the frequency should 

 follow more nearly the square of the dose since the complications men- 

 tioned will be relatively infrequent, whereas for a higher range of doses 

 the frequency should follow a power of the dose smaller than ^2. In cor- 

 respondence with this expectation it was found (Muller, 1940) that, at 

 doses between 375 and 1500 r, the translocation frequency appeared to be 

 nearly proportionate to the square of the dose. Although these data 

 were too few to be conclusive, they receive confirmation through the fact 

 that the same trend may be seen in the translocation frequencies reported 

 by Catsch, Kanellis, and Radu (1943) and Catsch (1948) for Drosophila, 

 when their results for the range 1000-2000 r are compared with those for 

 the range 2000-4000 r. Nevertheless, further data, on even lower doses, 

 are to be desired, especially since, as will be explained later, a change of 

 opposite sign in the relation (i.e., a reduction in the exponent) is expected 

 to set in at still lower doses than those yet tested for gross structural 

 changes in Drosophila. 



The microspores of the spiderwort Tradescantia have furnished the 

 main material for direct cytological observation of the structural changes 

 in chromosomes caused by ionizing radiation. It was shown by Sax 

 (1938, 1939, 1940, 1941) and confirmed by others (e.g., MarineUi et al, 

 1942; Thoday, 1942) that in this material the frequency of chromosome 

 fragments and bridges derived from single breakage varies linearly, i.e., in 

 direct proportion to the total X-ray dose. The same result was soon 

 afterward obtained by Marquardt (1941) on Bellevalia, by Carlson (1941) 

 on neuroblasts of the grasshopper Chortophaga, and by Reynolds (1941) 

 on the cheese mite Sciara. The frequency of gross structural changes such 

 as translocations and rings, representing more than one break, the ends 

 of which have exchanged, was, however, proportional to a power of the 

 dose higher than 1. When the doses were varied and yet the exposure 

 time was maintained constant (Sax, 1940, 1941) or so high an intensity 

 was used that the total exposure time was always very short (Sax and 

 Brumfield, 1943), this power was approximately the square. This result 

 does not disagree in principle from the Drosophila results because in this 

 material, mainly irradiated during interphase or early prophase, the 

 chromosomes remained relatively dispersed during the period between 



