L. H. GRAY 



extent to which the differences in biological effectiveness, which might have 

 been expected on the basis of spectra such as those of Figure 6, become 

 blurred when the S-rays are taken into account, but they are unsatisfactory 

 on account of an unavoidable arbitrariness in the definition of a S-ray. 

 Burch^o, for example, includes as S-rays all particles which have an initial 

 energy greater than 100 eV, which implies that clusters of three or more ion- 

 pairs are classified as high LET particles. It would seem to me that the 

 close association of more than three ion-pairs are needed to give rise to 

 chemical or biological effects which are characteristically different from those 

 produced by more widely spaced ion-pairs. 



Lea, from a study of the R.B.E. of different radiations with respect to the 

 production of chromosome structural damage in Tradescantia microspores 

 concluded that in order to produce a chromatid break a particle must 

 satisfy two criteria; it must have a LET at least as great as 6 ekV/[x, and a 

 range of at least • 1 [j.. These criteria ensure that a particle can cross a 

 chromatid thread and will leave at least • 6 ekV of energy in the material of 

 the chromosome in the course of its transit. In terms of Lea's estimates of 

 the range and i-ate of loss of energy by slow electrons, these criteria imply 

 that the effective electrons are those with an energy between about 1 • 5 and 

 2-8 ekV. These conditions are satisfied by the 'tail' of the track of every 

 electron of energy greater than 1 • 5 ekV, and in addition by a proportion of 

 the S-rays generated along the entire length of a fast electron track. A few 

 years ago I thought it might be of interest to calculate, for monochromatic 

 photon beams of different energy, the proportion of the total dose which 

 satisfies criteria of this kind. Details of such calculations are laborious and 

 only approximate. They have been published elsewhere^^ • -2. Figure 7 shows 

 the results obtained in one such calculation. The ordinate -q is the proportion 

 of the dose contributed by energy losses which satisfy the condition • 5 < () 

 < 3 • ekV, and is shown as a function of the primary photon energy. The 

 curves for hexane, water, and chloroform, differ on account of the increasing 

 proportion of the total energy which is contributed by photo-electrons from 

 the heavier elements oxygen and chlorine. The very pronounced hump in 

 the curves for water and hexane is an unexpected finding. In terms of the 

 extreme assumption that energy loss of the kind considered has unit efficiency 

 for a particular kind of chemical or biological damage [e.g. chromosome 

 breakage) and all radiations of lower LET have zero efficiency, the quantity 

 -q represents the apparent mean efficiency of monochromatic X radiation as 

 a function of photon energy. Most practical sources of X radiation have a 

 fairly broad spectrum, as shown at the foot oi Figure 7. It will be seen, how- 

 ever, that the peak of emission from a tube operated at 200 kV with an • 5 

 mm Cu filter, approximately coincides with the peak of the hump for water 

 and hexane. It might happen, therefore, that both lightly filtered and very 

 heavily filtered 200 kV radiation were less effective biologically than moder- 

 ately filtered radiation. These calculations were, in fact, prompted by a 

 report from Kirby-Smith and Daniels'^ that 250 kVp X radiation filtered 

 through 3 mm Al was 1 -33 times as effective as radiation generated at the 

 same kilovoltage but filtered through 4 mm Cu, in breaking Tradescantia 

 chromosomes. 



In conclusion I must add that all I have said regarding the distribution in 



11 



