MANNER OF PRODUCTION OF MUTATIONS 485 



breakage and union and hence did not have as much opportunity as in 

 Drosophila to form combinations more complicated than two-by-two 

 exchanges. Moreover, it was the aneucentric rather than the eucentric 

 combinations which were usually detected cytologically in Tradescantia, 

 unlike Drosophila, and the frequency of these aneucentric combinations is 

 expected to rise with dose more rapidly than that of the eucentric types. 



When, on the other hand, the dose applied to Tradescantia was varied 

 by altering its duration, with the longer treatments lasting several min- 

 utes or tens of minutes, the frequency of exchanges was observed by Sax 

 to vary according to approximately the % power of the dose as it had in 

 the Drosophila work, although for a different reason. With the dose 

 varied by delivering different numbers of widely separated fractional 

 exposures, a linear relation was obtained (Nebel et al., 1942). Accord- 

 ing to Sax the reason for such effects of protracted treatments is that the 

 broken ends, in the given interphase stages, are able to undergo union 

 before the irradiation is finished so that their restitution (and, in some 

 stages, isochromatid union) often occurs before other broken ends have 

 been produced which might otherwise, with a shorter dose of the same 

 total amount, have undergone exchange unions with them. This time 

 restriction on joining tends to divide up the prolonged or fractionated 

 dose into a number of more or less self-contained (even if overlapping) 

 smaller fractions ; the broken ends derived from each fraction are only to 

 a small extent available to those derived from other fractions. Thus the 

 total frequency of exchanges formed tends to be more or less linearly pro- 

 portional to the number of these fractions given, although within each 

 fraction the frequency varies as the square of the dose deUvered in that 

 time. 



Results of fractionating the treatment, with different intervals between, 

 indicated that most of the unions took place within 10-20 minutes after 

 breakage, although a relatively few broken ends remained open and join- 

 able for a very much longer period. As mentioned in Chap. 7, however, 

 Lane (1951), on the basis of his observations, concludes that in the Trades- 

 cantia material the ends remain open until much later, much as in sperma- 

 tozoa, and that the lesser effectiveness of the protracted and fractionated 

 treatments is caused by the irradiation inducing, in addition to the break- 

 ages, a lowered susceptibility of the chromosomes to breakage by subse- 

 quently applied radiation. Recent results by Swanson (1949) show that, 

 even 18 hours after exposure to X rays, the frequency of exchange unions 

 in Tradescantia microspores can still be increased substantially by appli- 

 cation of infrared. Unless this is (as Swanson believes probable) due to 

 the production of new breaks, it shows that many breaks, due otherwise 

 to restitute, have remained open for this length of time. Curiously, in 

 barley seeds, Kaplan (1951) finds that fractionation causes an increase in 

 chromosome aberrations. After much computation, he concludes from 



