THE AMERICAN ORIGIN OF AGRICULTURE. 503 



That the varieties used for this purpose are as old or older than 

 those grown for fruit is indicated by the fact that, like the sweet potato, 

 taro and sugar-cane, they seldom produce flowers. Furthermore, among 

 all savage tribes the varieties valued by civilized peoples as fruits are 

 relatively little used, far greater popularity being enjoyed by the so- 

 called ' plantains, ' not edible in the raw state, even when ripe, though 

 nearly always cooked and eaten while still immature, or before the 

 starch has changed to sugar. They are also in many countries dried 

 and made into a meal or flour often compared to arrowroot. 



In dietary and culinary senses the breadfruit also is as much a 

 vegetable as the taro or the sweet potato ; as a fruit it would be no more 

 likely to be domesticated than its distant relative the Osage orange. 

 The farinaceous character of the breadfruit also probably explains its 

 relatively greater importance among the Polynesians than in its orig- 

 inal Malayan home, as shown by the propagation of numerous seedless 

 varieties. The popularity of the breadfruit among the Polynesians was 

 further extended by the discovery that the fruits could be stored in 

 covered pits, the prototypes of the modern silo. 



If the domestication of the banana is to be ascribed to cultivators 

 of root-crops, the same reasoning applies with even greater propriety 

 to cereals. Tribes accustomed to subsist upon mangoes, dates, figs or 

 similar fruits which require no grating, grinding or cooking, and are 

 eaten alone and not with meat, would not develop the food habits and 

 culinary arts necessary to equip primitive man for utilizing the cereals. 



Wild bananas and their botanical relatives are natives of the rocky 

 slopes of mountainous regions of the moist tropics where shrubs and 

 trees prevent the growth of ordinary herbaceous vegetation. The com- 

 mencement of the culture of cereals by fruit-eating natives of such 

 forest-covered regions is obviously improbable, but such an undertaking 

 would be a comparatively easy transition for the meal-eating cultivators 

 of root-crops, since the grasses and other plants domesticated for their 

 seeds are exactly those which flourish in cleared ground and are prompt 

 to take advantage of the cultural efforts intended for other crops. Thus, 

 the Japanese have by selection secured a useful cereal from the common 

 barnyard-grass (Panicum crus-galli) just as they have made a root- 

 crop of the burdock. Accordingly, we should look to some taro-growing 

 tribe of southeastern Asia as the probable domesticators of rice, and to 

 other cultivators of root-crops for similar services in taking up the 

 somewhat less tropical cereals, sesame and Guinea corn. That root- 

 crops preceded cereals in America was inferred above partly from the 



placed Heliconia in cultivation among the Polynesians. In the time of Oviedo 

 the natives of the West Indies made hats, mats, baskets and thatch from the 

 leaves of Heliconia, and the starchy rootstocks were eaten. 



