CHROMOSOME ABERRATION 



least two electrons is necessary to provide a disturbance adequate to initiate an 

 exchange. I prefer the alternative view, which I think is supported by the dose 

 relations observed with neutrons and a-particles. At not too high doses, exchanges 

 which follow a square-law dose with X-rays are linearly related to dose when 

 produced by neutrons, but even so show a small square-law component at high 

 neutron doses. Again, they are linearly related to dose with all doses of a-radiation 

 so far studied. Having regard to the number of electrons, protons and a-particles 

 by which a cell is traversed when exposed to a given dose of X-, neutron- or a- 

 radiation, the dose relations observed with these three radiations are just those to 

 be expected if exchanges result from two independently produced breaks. More- 

 over, in the case of a-radiation, such low doses have been used that most of the 

 aberrations observed have been produced in cells whose nuclei have been traversed 

 by only one a-particle. It would seem difficult to account for the high efficiency 

 with which a-particles produce aberrations if the induction of aberrations is restricted 

 to occasions on which a particle passes through or in the immediate vicinity of a 

 region in which two chromatid threads are lying in contact. 



We all exercise some degree of selection, conforming our hypothesis to fit the facts 

 which impress us most, and it may be that I am placing too much emphasis on the 

 quantitative aspects of the dose relations observed with different types of radiation. 

 The fact that Revell has adopted the ' contact first ' hypothesis for chromosome 

 breakage cannot but result in a critical re-examination of the ionizing radiation 

 data, and will, we hope, lead to further experiments designed to decide between 

 the alternative hypotheses. 



R.B.E. FOR X-RAYS OF DIFFERENT TUBE VOLTAGES 



In connection with the papers by Kirby-Smith and Daniels, and by Swanson, I 

 would like to add a few remarks concerning the influence of kilovoltage and filtra- 

 tion on the yield of aberrations produced per unit dose. On a number of occasions 

 (Lea\ Gray"^, Spiers^) attention has been drawn to the fact that the mean energy 

 of the secondary electrons generated in tissue remains almost constant over the entire 

 range of photon energies from 25kV to 100 kV, and that on this account it would 

 be surprising if biological efficiency showed any appreciable dependence on X-ray 

 tube kilovoltage over the range 50-200 kV. As is well known, the constancy of the 

 mean electron energy arises from a fortuitous balance between the varying pro- 

 portions of the more energetic photoelectrons and the low energy recoil electrons, 

 and I thought it would be of interest to see how far this balance might be disturbed 

 in the case of a phenomenon such as the production of chromosome structural 

 damage for which we have reason to believe that the contribution of the very slow 

 electrons should be heavily weighted. From a consideration of the relative effective- 

 ness with which X-rays, neutrons and a-particles produce chromosome aberrations 

 in Tradescantia microspores, Lea concluded that in order to break a chromatid 

 thread an electron must [a] have sufficient residual range (0 • 1 [x) to traverse the 

 thread and {b) dissipate at least 0-5 keV of energy in crossing the thread. (Lea^, 

 p. 276). In terms of the energy of the electron as it enters the chromatid thread, 

 this limits chromatid breakage to electrons having energies between about 1 -6 keV 

 and 2 • 8 keV. Having applications to radiation chemistry as well as to chromosome 

 breakage in mind at the time I made my calculations, I have assigned unit efficiency 

 either to 



(fl) electrons having energy of 0-5keV (Column I of the Table) ; or 

 {b) electrons having energy between 0-5 keV and 3 keV. (Column II of the 

 Table) 

 and zero efficiency to all contributions to the total dose from electrons whose energies 

 lie outside the assigned limits. 



278 



