498 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1962 



annual forms have been bred and brought into use only in historic 

 times. For a study of the development of the cottons, therefore, the 

 primitive perennials, both wild and cultivated, are the natural start- 

 ing-point. Primitive perennial types of the cultivated cottons are not 

 difficult to find. The modern textile industry, consisting of great 

 mills dependent for supplies on the cultivation of fairly uniform types 

 over large areas, is of very recent origin, and in all the older cotton- 

 growing countries the remnants of the old textile crafts are still to be 

 found. The craftsman — spinning, weaving, and dyeing the cloth in 

 traditional ways and on a small scale — draws his raw material from 

 small plots, or occasional plants grown in gardens or house com- 

 pounds. It was from these relics of village crafts that the primitive 

 forms of the world's cottons were collected. They are to be fomid 

 over wide areas in India and southeast Asia, in the Sudan and West 

 Africa, and in the West Indies, Central America, and northern South 

 America. They occur less commonly in the Pacific islands and even 

 in northern Australia. 



These cottons readily run wild and establish themselves in waste 

 places and hedgerows, and on the margins of cultivated land. Watt 

 and other early writers regarded many of them as truly wild, and 

 ancestral to the cultivated cottons. In collecting, it became apparent 

 that there was no line of demarcation between the wild plants of waste 

 places; the occasional plants of house compounds and gardens that 

 grew from carelessly scattered seeds but which were regularly har- 

 vested; and the true crop plants of neighboring fields. It then be- 

 came important to decide whether the wild plant was the original from 

 which the cultivated form was developed, or whether it arose as an 

 escape into natural or seminatural vegetation from cultivated lands. 



As our knowledge increased, so the number of forms that could be 

 regarded as independent of human influence diminished. A. Cheva- 

 lier [4] stated that, of the cottons now to be found in West Africa, 

 none occurs in undisturbed vegetation, and all show signs of having 

 reached their present habitat as cultivated plants. In the Sudan 

 and Ethiopia [21], it is also clear that the cottons exist in association 

 with man, and not wild in natural vegetation. Moreover, in no 

 locality in India are the uncultivated forms of cotton to be found 

 growing in anything like natural vegetation. They have, in fact, all 

 the characteristics to be expected of the abandoned relics of the crops 

 and crafts of an earlier culture. 



The situation is similar in the New World. S. G. Stephens [17], 

 collecting cottons in Central America, found that it was useless to 

 look for material anywhere but in villages and on village lands. In 

 the West Indian islands also, the cottons are to be found growing as 

 commensal plants, preserved and harvested for domestic use even if 

 not deliberately sown. There are a few reports of New World cottons 



