376 



Tn ,i;'\ia,iiiii,i,' what (le.fireo of iiuportaiKO. is to ho assi^-nod to 

 each of the ahdvc. the (iist consideration is tliat man has sek^ted 

 the (ireater Vani Inmi soniethini;' jirowinii' wihl. and therefore we 

 must ascertain as far as ])ossil:)le wliat ajipearanee the wild |)arent 

 liad, and wliat characters tlie first cultivators would he likely to con- 

 sider. Xow the Greater Yam has several close allies amonor the 

 wild yams of Eastern Tropical Asia : and if not from these, from 

 something very similar it must have taken its origin. These allies 

 grow in moist regions where forest prevails, finding their oppor- 

 tnnities for vigoronsly fruiting where the canopy of overhead foliage 

 l)as heen disturhed, and the sunlight let in. They have tuhers 

 which are hnried deeply in the soil ont of the reach of heing 

 uprooted hy wild pigs : and these tubers are white fleshed and 

 esculent, having none of the poisonous principle which is present 

 in the tul:)ers of those more remote species of Dioscorea which 

 are surface rooting. The tuhers are thin, thickening a little down- 

 wards, but not greatly, and they mav liranch a very little, or 

 not at all. Jungle tribes eat the wild vams of this type: but 

 it is not much that they return for the labour of digging them 

 out. Onc-e the writer engaged a Avoman of a jungle tribe to dig 

 \\ild yams for him from a forest in Xorthern Burma where 

 tliey are particularly common, and obtained less than a meal for 

 two persons as the result of four hours work. But Avithout doul)t 

 from wild yams yielding such a meagre return, junirle man 

 commenced the selection of the Greater Yam. First the tril)es 

 would eat what they could find : then becoming more settled in 

 habitation tliey would find it necessary to encourage the plants 

 to groM- about the settlements, and would protect them a little: 

 thirdlv tliey would plant them in the immediate neighbourhood of 

 the villages: and out of this procedure thev would evolve a regular 

 cultivation, selecting naturally the parents for projiagation wliicli 

 were most jiroductive. The second stage of development 1)eing 

 reached in this way. it Avould he realised advantageous to grow 

 plants Avith abbreviated tuhers so that the tilling might be less 

 laborious and tlie digging out more easily accomplished. There- 

 fore, second to yield, man Avould operate on the length of the root, 

 which in tliis stage of social development he could now protect 

 against uprooting by wild pigs — at least in some measure. 



It is perfectly certain that he Avould not select for excellence 

 in flavour at this stage, for the jungle man everyAvhere eats ap- 

 parently by no means unwillingly, the most nauseous vegetables. 

 Colour he might select for, as a toy which pleases, long before think- 

 ing of palatahility. Shape of leaf, rootiness of the tuber, upgroAA'- 

 ing fleshy shoots, and other sucli characters, he Avould not notice. 

 All these variable features, then, may be put doAvn as secondary to 

 tlie length of the tuber; and in that feature the first division in the 

 classification of the races will be based. 



