458 Allopolyploids 



somes in these species shows that S. Townsendii has 126, and the 

 other two species have respectively 70 and 56. To reconstruct 

 the probable sequence of events, a 35-chromosome gamete of S. 

 altemiflora united with a 28-chromosome gamete of *S. stricta to 

 form a hybrid which had 63 chromosomes and was probably 

 highly sterile. In some way that we cannot explain, the number 

 of chromosomes of the hybrid became doubled to produce a plant 

 with 126 chromosomes, which was practically fully fertile. This 

 plant was *S. Townsendii. This species has four genomes, but 

 the two genomes of S. altemiflora are so different from those of 

 S. stricta that chromosomes from the two species never pair with 

 one another and, as a result, only bivalents are formed. Each 

 bivalent consists of either two altemiflora chromosomes or two 

 homologues of stricta. 



Numerous similar amphidiploids that arose from crosses be- 

 tween two species of the same genus might be cited. To men- 

 tion a very few, there are Primula kewensis, which arose from 

 a cross between P. floribunda and P. verticillata; Digitalis mer- 

 tonensis, which arose from crossing among themselves hybrids 

 between D. purpurea and -D. ambigua; and Galeopsis Tetrahit, 

 which arose in nature from a cross between G. pubescens and G. 

 speciosa and was also synthesized artificially by Miintzing, who 

 crossed these species together. Apparently a large number of 

 cultivated plants have originated as allopolyploids. Some species 

 of cotton have 13 pairs of chromosomes; others have 26 pairs. 

 The former group includes wild species from Asia, Africa, Aus- 

 tralia, Central America, and the Galapagos Islands and the cul- 

 tivated Asiatic cottons, whereas the 26-chromosome group in- 

 cludes wild plants from Mexico, South America, the Galapagos 

 Islands, and the Hawaiian Islands and the American types of 

 cultivated cotton. Skovsted has suggested that the 26-chromo- 

 some American varieties are amphidiploids which arose from a 

 cross between 13-chromosome types from America and Asia. 

 This suggestion has been strongly supported by Beasley's ex- 

 perimental evidence, based on a resynthesis of a similar 26- 

 chromosome type from a cross between an American 13-chromo- 

 some cotton and a type from Asia which also had 13 chromo- 

 somes. Apparently our cultivated tobacco arose as a 48-chromo- 

 some amphidiploid between twb 24-chromosome species. Al- 



