534 RADIATION BIOLOGY 



but somehow counteracts the process by which the broken ends join into 

 new combinations. It appears to do this, as Kaufmann and HoUaender 

 pointed out, by favoring restitution rather than terminal deletion or 

 isochromosome formation, since they found that the ultraviolet given in 

 addition to the X rays did not raise the frequency of dominant lethals 

 beyond that expected as an additive function of the two treatments. A 

 greater increase than this in dominant lethals would have resulted if some 

 of the X-ray cases, which without ultraviolet would have become trans- 

 locations, inversions, or small deletions, had been caused by the ultra- 

 violet to have their broken ends remain unconnected or to undergo iso- 

 chromatid union. 



A possible basis for such increased restitution was provided by Swanson 

 (1942, 1944) in his observation that the ultraviolet-treated Tradescantia 

 chromosomes underwent a shortening, and a change in their appearance 

 suggesting a stiffening of their matrix. It is true that this would not 

 readily explain the prevalence of terminal deletions as opposed to inter- 

 changes which was found in some of the ultraviolet-treated plant mate- 

 rial. Nevertheless, the influence of the ultraviolet on the X-ray breaks 

 shows that, in view of this restitutional effect, the conclusion cannot yet 

 be drawn that ultraviolet produces fewer primary breaks than X rays, 

 when doses of equal potency in causing point mutations are compared. 



The yield of breaks which had remained ununited until the time of 

 cytological observation was found by Swanson (1940, 1942) in his 

 Tradescantia material to show a straight-line relation to the dose of 

 ultraviolet used, over a range of doses, the extremes of which differed by a 

 factor of about 8. This observation is of great interest in that it indicates 

 that these breaks are produced by individual singly-acting activations. 

 With ultraviolet, since no tracks are produced, the number of activations 

 within any given volume varies with the dose. Thus, if activated 

 radicals cooperate in producing the changes, these would vary as a power 

 of the dose received during any fixed exposure time (this power being 

 equal to the number of cooperating particles). The frequency-dosage 

 curve would in that case rise concavely as viewed from above. Unfortu- 

 nately, in these experiments the dose was changed only by varying the 

 time of exposure, whereas evidence on this question can be clear only if the 

 frequency produced is shown to be independent of the timing of the 

 exposures given. This was probably the case in this work, however, 

 if it is permissible to judge by the results obtained for point muta- 

 tions in Drosophila pole cells by the Altenburgs and referred to on 

 p. 538. 



On the other hand, Stadler and Uber (1942) found a convexly rising 

 curve instead of a straight line in maize, when they used chromosome 

 deficiencies phenotypically evident in the endosperm as a measure. 

 However, they calculated that the deviation from linearity was prob- 



