Chromosomes and Genes 147 



inherited. Dt then acts exactly like a mutator, confined to a single 

 locus. No rearrangements could be detected, but the Dt locus is near 

 or within the end knob, which possibly is heterochromatic. Of course, 

 rearrangements in maize must be, as we emphasized earlier, minimally 

 of ten times the order of magnitude as in Drosophila salivary chro- 

 mosomes in order to be detectable. In this and all other comparable 

 cases the facts did not lead beyond these statements, based upon 

 Rhoades' work. 



It stands to reason that, in spite of the special features mentioned, 

 these somatic mutations are genetically not different from standard 

 mutations. Demerec expressed this by saying that we have all ranges 

 of mutation frequency, rare mutations, frequent ones, very frequent 

 ones in the cases of mutators studied above, and finally these extreme 

 frequencies. His statement is certainly formally correct, but only when 

 we give it a completely different meaning from the one intended. For 

 Demerec the statement meant that only ordinary gene mutations are 

 involved, but at different rates. To us it means that mutations are 

 position effects, since the proof has been provided by McClintock 

 that "mutable genes" are a special type of position effect caused by 

 frequent rearrangements. 



At this point, therefore, McChntock's work (see 1951) on similar 

 phenomena is applicable to the present problem. It is very difficult to 

 understand in its original terminology, which was chosen before the 

 real meaning came to light. We shall present it so that the relation 

 to the work on position effect in Drosophila is made clear by the use 

 of the Drosophila terminology also for McClintock's work. We re- 

 member that the chromosomal breakage-fusion-bridge cycle in maize 

 induced the rearrangements in other chromosomes. It also accounts 

 for the origin of new mutable loci of the type discussed for the earlier 

 maize work. This means that in plants which were the seff-pollinated 

 progeny of those which, in early development, had undergone that 

 breakage cycle, a "burst" of new mutable loci appeared. It may help 

 us to understand the complicated facts if we say that this breakage 

 cycle (i.e., basically, the breakage) acts as a primary mutator. 



The next fact is that, in the unstable line derived from the parents 

 with "mutator," a break at a definite point of the chromosome appears 

 with definite frequencies and at definite times, a break which we may 

 call a secondary mutator. This break, obviously one of the direct 

 actions of the primary mutator (as defined above, i.e., a consequence 

 of the chromosomal abnormalities of the fusion-bridge cycle), is un- 

 fortunately described as a dominant locus, which then behaves hke 



