422 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1961 



tion, these all helped to carry conviction. Even in Oenothera the 

 occurrence of chiasmata, the inference of crossing-over, could be used 

 to explain the modes of inheritance and the origins and kinds of muta- 

 tions [8]. 



The solution of the problem of crossing-over was, however, ob- 

 structed by another, to me, unexpected but inseparable development. 

 The study I had made of meiosis, and my interpretation of what a 

 hundred others had seen in a great variety of plants and animals, made 

 it clear that the chiasma had consequences which were as important as 

 its causes [9]. People had supposed that the chromosomes were 

 paired at the first metaphase of meiosis because they were, in a mysteri- 

 ous sense, attracted to one another. This was, they thought, the climax 

 of the sexual process. What could be more natural ? But I had 

 found that, after pairing as single threads, the chromosomes fall apart 

 as double threads. They do not attract, they repel one another after 

 they become double; and they are held together only by chiasmata, 

 by the exchanges of partner between their chromatids, after crossing- 

 over. When pairs of chromosomes fail to form chiasmata, whatever 

 the cause, the consequence is that they lie on the spindle, unconnected 

 and unorientated. They then fail to pass to opposite poles. With- 

 out chiasmata meiosis itself, with Mendelian segregation, the reduction 

 of chromosome number and the alternation of haploid and diploid 

 in the sexual cycle, all these fail to ensue. 



Thus there was, in my view, a universal causal sequence : crossing- 

 over— 5>chiasmata—>chromosome pairing^segregation and reduc- 

 tion—asexual reproduction. 



This reversal of the mechanical interpretation of meiosis made it 

 possible to describe meiosis in the same physicochemical terms as 

 mitosis. But what mattered first were its genetic implications. The 

 new principle seemed to be true of all plants and animals (except 

 male Droso'pliila). It therefore meant that meiosis and crossing- 

 over had come in together, at one step. Crossing-over from its origin 

 must have been coextensive with sexual reproduction. What Morgan 

 had hoped to imply I was now forced to assert. The reason why the 

 chromosomes were divisible into units or genes was that everywhere 

 their division into such units was a condition of meiosis and hence of 

 sexual reproduction. This, of course, made sense in terms of selec- 

 tion, adaptation, and the evolution of sexual reproduction itself. For 

 if the chromosome were not divisible into genes, if it were not capable 

 of crossing-over, it would be inherited as a block and no genes could be 

 revealed either to the geneticist by his experiments or to nature by 

 her selection. The origmal system would never have survived. 



