Chromosomes and Genes 145 



1939). I still think it possible that it was not a real error but a rare 

 eflFect working only when some previous condition was present in 

 some chromosome. 



At this point attention should be drawn to a little explored field. 

 A number of records at hand, none of them studied very thoroughly, 

 indicate that a causal relation between simultaneous or consecutive 

 different steps of mutation may exist. Skovstedt (1943) has described 

 such a situation in Nadsonia, and I have found vesy suggestive 

 evidence in Drosophila ( Goldschmidt, 1947). In both it appeared that 

 the occurring of one mutation increased the probability of another 

 occurring. In Drosophila cases have been described in which two 

 mutations in di£Ferent chromosomes appeared repeatedly together. 

 Though a completely convincing analysis is lacking, the observations 

 should be kept in mind as possible examples of the mutator type, 

 better proof of which may be found one day. 



However, there is available in maize very good cytological in- 

 formation (McClintock, 1941-1951; see 1951), which shows how a 

 chromosome breakage without rearrangement of the usual type can 

 produce other chromosomal changes; we might call this a mutator 

 action. McClintock found what she calls the breakage-fusion-bridge 

 cycle. There are two types of it. One is called the chromatid type of 

 the cycle. It is found when a chromosome with a newly broken end 

 enters the telophase of meiosis. Fusion occurs between two sister 

 chromatids at the position of previous anaphase breakage, resulting 

 in bridge formation in the next anaphase. This may continue over 

 many mitotic cycles. If, however, a chromosome with a newly broken 

 end is introduced into the zygote by each gametic nucleus, the broken 

 ends of the two chromosomes — not chromatids — may fuse. This estab- 

 lishes a dicentric chromosome and a diflFerent type of breakage-fusion- 

 bridge cycle is initiated. In the telophase nuclei the fusions now occur 

 between the broken ends of chromosomes. The important point for 

 the present discussion is that this cycle may initiate breakage events 

 in other chromosomes and also the origin of mutable loci. In the 

 terminology used before, this breakage cycle acts as a mutator. (We 

 shall discuss these facts in more detail in the next section.) 



ccc. Mutable loci in maize. — The most complete and interesting 

 contribution to the present discussion comes from McClintock's work 

 on mutable loci in maize (see 1951). It shows that maize, the most 

 favorable genetic material in plants, second only to Drosophila, 

 supplies the same type of information on position effect; hence we 

 can no longer doubt that it is a general phenomenon. Simultaneously, 



