510 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1962 



Islands. Species are now known also from continental South America, 

 Arabia, and Africa ; they are listed and described by J. H. Saunders 

 [22]. The wild species do not bear seed hairs that can be spun, and 

 one of the major problems of the origin of the cottons is to account 

 for the origin of lint. The simplest hypothesis is that the four lint- 

 bearing species derived their lint from one common ancestor, and the 

 history of the true cottons will now be reconstructed on this basis. 

 The chromosome complement in Gossyplum is basically n=13; all 

 the wild species save one are diploid. The two Old World species of 

 the true cottons are also diploid. The two New World species of 

 cotton and one wild species [G. tomentosum in Hawaii) are tetraploid. 

 Of the diploid wild species, the African species of the section Anomala 

 are comparatively closely related to the Old World diploid cottons. 

 There are two species, G. anomalmn and G. tiiphyllum^ and both 

 have been crossed with the two Old World cottons. The hybrids are 

 moderately fertile, and gene transfer and genetic analysis have been 

 carried out successfully [23]. D. U. Gerstel [9] has shown that the 

 chromosome complement of G. herhaceum is very closely homologous 

 with that of G. anomaluin. The complement of G. arhoreum differs 

 from those of G. herhaceum and G. anomahim by one translocation. 

 It is in G. herbaceum^ therefore, that one would logically look for the 

 wild prototype of the diploid cottons. The Anomala are distributed 

 in arid areas on the southern borders of the Sahara and in Angola 

 and southwest Africa. The wild form of G. herhaceum^ race a/r/- 

 canum, occurs in natural vegetation right across the dry bushveld and 

 semiarid tracts of southern Africa from Mozambique to Angola. It 

 occupies, in fact, an area adjoining that of the most closely related 

 wild lintless species, but it has no obvious contact with the cultivated 

 races of its own species. It has all the characters that would be ex- 

 pected of the original wild form: the perennial bushy habit and 

 the hard seeds with a marked period of dormancy adapted to survival 

 in the wild, and a coat of rough, wiry lint on the seeds, long enough 

 and copious enough to be spun if nothing better offered, but of the 

 lowest quality in terms of modern textile materials. The only sub- 

 stantial difficulty in regarding it as ancestral to the cultivated cottons 

 is the enormous distance between the natural habitat of africanum and 

 the areas occupied by the primitive cultivated form of the species, 

 race acerifolium^ in Ethiopia and southern Arabia. If africanum is 

 ancestral to the cultivated forms, it must have been carried far to 

 the north at a very early date, or its range must formerly have been 

 vastly greater than it is now. Travel up and down the East African 

 coast has gone on from ancient times, and the trade in gold from the 

 area, in what is now Southern Rhodesia, in which africanum grows 

 wild, is of unlalo^vn antiquity. It seems most likely, therefore, that 



