430 BREAKDOWN OF GENETIC SYSTEMS 



rule, only the workable type of chromosome with two ends and one 

 centromere. The changes in its size may have arisen in several 

 ways, but losses will often have been eliminated with cell-selection 

 by lower viabiHty, and inversions, unless including the centromere, 

 will not be visible. Observation of mitosis to show changes in 

 cells derived from irradiation is therefore restricted to what have 

 been called " translocations." These have been assumed by 

 Lewitsky and Sizova (1934) always to be simple " donations " from 

 one chromosome to another. Study of meiosis in the progeny of 

 X-rayed plants, however, shows a great number of interchanges. 

 Large numbers have been identified in Zea, occurring at random, 

 but necessarily balanced since they have survived (Anderson and 

 Clokey, 1934). In all the products of structural change in maize, 

 there is no case of a simple fusion or translocation, or indeed of 

 any change involving the attachment of unbroken ends of chromo- 

 somes, although there is no mechanical reason why such changes 

 should not survive (McCHntock, 1931 ; cf. Ch. XII). Changes of 

 size must therefore often be due to the interchange of unequal 

 segments, such as has been identified in Zea (Burnham, 1932). 



On the assumption of a simple donation in Crepis capillaris, 

 Lewitsky and Sizova have, however, concluded that since long arms 

 *' give " to short arms much more often than they receive from them, 

 there must be a specificity in the direction of donation. But when 

 we assume interchange we see that this apparent specificity is 

 nothing more than the inevitable result of a size difference. When 

 a large body exchanges parts at random with a small body it 

 stands to lose by the transaction. There is therefore no specificity 

 in the direction of transfer. 



With regard to a possible specificity in the places of breakage, 

 some workers have concluded on occasional evidence that breakage 

 is more frequent near the centromere than elsewhere, but this is not 

 generally so. It may be, however, that the X chromosome is more 

 frequently broken than the autosomes in Orthoptera (White, 1935). 



Not only may the genes and gene-string be broken by irradiation, 

 but also, as we have seen, the centrosome may be prevented from 

 dividing. We do not know whether this is a direct or indirect effect. 

 It is possible, however, that a direct change may be induced in the 



