NATURE OF THE GENETIC EFFECTS 365 



bridge between the daughter mielei, and this bridge may later, through 

 a mechanism not yet well understood, lead to the death of the cells 

 involved. This has been shown to happen even when their genetic 

 deficiency is compensated for by supplying them with an additional 

 chromosome of appropriate type (Pontecorvo and Muller, 1941; Muller 

 and Pontecorvo, 1942b; Pontecorvo, 1942). 



Sometimes, instead of failing to enter the nuclei, or forming a fatal 

 bridge, the isochromosome stretched in the bridge becomes broken again, 

 and one or both of the two fragments may then be pulled into their 

 respective nuclei. Thereafter, their own daughter chromatids, because 

 of their broken ends, repeat the story of dicentric isochromosome and 

 bridge formation. This process is in the animal material studied unlikely 

 to go through many cycles without the affected chromosome finally 

 becoming lost from the descendant cells, or else killing them by bridge 

 formation. In some material, e.g., maize, the above "breakage-fusion- 

 bridge cycle" may be repeated almost indefinitely, as McClintock (1932, 

 1938a, 1939 et seq.) has shown. In this case, since the chromosome is 

 broken anew at every mitosis, and since each new break is likely to be in 

 a different position than before, the genetic composition of the repeatedly 

 patched remainder becomes more and more abnormal, as some chromo- 

 some parts are lost while others become increasingly reduplicated. 

 Accordingly, the genetic composition of the descendant cells comes to 

 depart ever further from the normal, to their increasing detriment. 



Thus, when breakage of a chromosome is followed by union between 

 identical "sister" (mother and daughter) fragments to form isochromo- 

 somes, genetic deficiencies and sometimes other genetic abnormalities 

 inevitably follow, by one means or another. If this happens in the germ 

 track of the main or sporophyte generation of a higher plant belonging to 

 an ordinary diploid species (having two complete sets of chromosomes in 

 that generation), and if the affected cells manage to survive until the 

 haploid or gametophyte generation (that with one set of chromosomes), 

 those with the deficiency are then killed off in the latter stage. At least 

 this is true in the male gametophyte, since this metabolizes more com- 

 pletely on its own account than the female gametophyte does and hence 

 has more use for its genes. For now this tissue no longer has a second, 

 normal set of chromosomes to mitigate the effect of its genetic abnor- 

 mality. In this way the gametophyte generation (at least that of the 

 male, and to a lesser extent that of the female) serves as a sieve to weed 

 out such cases. 



In animals the corresponding haploid stage, found in the gametes, does 

 not perform a hke selective function, since the limited type of metab- 

 olism of these cells does not depend upon their genes, which are in a 

 dormant state at that time, but upon the products of the genes present 

 before reduction, and of the genes in diploid supporting (nurse and 



