THE POSITION EFFECT 



of their products, since diffusion camiot be instantaneous. Ephrussi, 

 however, has shown that the distances are too great for this 

 explanation to serve ail cases. 



The unit of physiological action as shown by the position effect 

 can be compared with the unit of X-ray breakage but not with 

 the unit of crossing-over. Thus the scute gene o( Drosophila mclano- 

 gaster undergoes many natural and induced mutations. They all 

 affect the number of bristles on the thorax and head. They are all 

 due to rearrangements (as Muller and Prokofieva found) in a series 

 of six chromomeres or polytene bands at the distal end of the 

 X chromosome. The rearrangements involve inversion, sometimes 

 of the greater part of the chromosome, and sometimes of segments 

 within the gene. The distinction between gene mutation and 

 structural change is thus arbitrary in the last resort. The whole 

 set of six segments being concerned with bristle development, the 

 two ends of the gene cannot be marked and crossing-over between 

 them cannot therefore be detected. It is unhkely to occur, however, 

 since none is found between scute and its neighbour yellow except 

 in structurally hybrid or triploid flies. Thus, in the scute complex, 

 the unit of crossing-over is larger than the unit of action, and this 

 in turn is larger than the unit of breakage and of self-reproduction, 

 which last corresponds to the units of visible discontinuity in the 

 chromosomes. 



On the view that the genes are integrated in themselves the 

 breakage of the chromosome should be inherently liable to disin- 

 tegrate them and so upset, for some distance on either side of the 

 break, the joint and adaptive action of the genes and their parts. 

 Like most other mutations, position effects are, in fact, nearly 

 always deleterious and indeed often lethal in the homozygous state. 

 They should, therefore, usually be lethal to the haploid gametophyte 

 and it is perhaps for this reason that position effects have rarely been 

 found in plants: cells having breakages that would yield position 

 effects presumably die. 



The disintegrating effect of chromosome breakage on complex 

 genes is reflected perhaps in the cytological effects of breakage. The 

 normal nucleic acid attachment cannot be correctly arranged at the 

 mitosis following breakage by irradiation (or by misdivision of 

 the centromere). Consequently, in a proportion of cells, the repro- 



331 



