THK ORIGIN OF SPECIES 



tetraploid. Generally, no such confusion exists, yet the distinction between 

 these two types of polyploidy is perhaps always one of degree rather than 

 an absolute distinction. By definition, tlie several haploid sets which com- 

 prise the chromosome complement of an autopolyploid do not differ from 

 one another in any greater degree than do the two haploid sets of the 

 corresponding diploid. But the first requirement for the formation of an 

 allopolyploid is that the two parent species must be able to form a viable 

 (though not necessarily fertile) hybrid. As discussed in the preceding 

 chapter, this becomes less and less probable as the relationship of the 

 potential parent species becomes more remote. The most usual situation 

 is that both parental species belong to the same genus, although many 

 cases are known in which allopolyploids have been formed between dif- 

 ferent genera of a single family. Most usually, the relationship is close 

 enough so there can be no doubt that there is a significant degree of 

 homology between the chromosomes of the parental species. Frequently, 

 this may be evident in terms of a limited amount of synapsis in the Fi 

 hybrid. Thus it seems probable that some degree of homology between 

 the several genomes of a polyploid is necessary for its formation. In auto- 

 polyploids, this homology is substantially complete, while in allopolyploids 

 it is markedly incomplete. 



AUTOTETRAPLOIDY IN EVOLUTION 



By the use of colchicine, a large number of experimental autotetraploids 

 has been formed and investigated, but autotetraploids are also known 

 from nature. Miintzing reviewed fifty-eight well authenticated examples 

 in 1936 and expressed the opinion that a list of over one hundred cases 

 could have been compiled. The study of polyploidy in all of its phases was 

 greatly stimulated by the introduction of the colchicine technique, and 

 there can be no doubt that the list would be much longer if a compre- 

 hensive review were to be published now. Autopolyploidy does not seem 

 to lead to the formation of new species, but only to well-marked varieties. 

 Yet there is considerable reproductive isolation between a diploid and its 

 autotetraploid, because the hybrid between them is a triploid ( three hap- 

 loid genomes in each somatic cell). Triploids are highly sterile because 

 the distribution of the chromosomes at meiosis is irregular. A few of 

 Miintzing's cases may be discussed below. 



Phleiim alpinum (a near relative of Timothy) is a grass which has a 

 diploid race with 14 chromosomes and a tetraploid race with 28 chromo- 

 somes. The two are morphologically distinct, and arc partially separated 

 geographically. The diploid race is found only in Scotland, while the tetra- 

 ploid race is found both in Scotland and in northern Scandinavia. It is 

 quite characteristic that tetraploids range more widely than their diploid 

 ancestors, frequently invading territories which are geologically more 

 recent. Crosses between the two races are successful, but the Fi is sterile, 

 being triploid. The case of Nasturtium officinale is similar. Races with 32, 

 48, and 64 chromosomes are known, and are respectively diploid, triploid, 

 and tetraploid. As in the case of Phleiim, the polyploid races have a more 



296 



