GENEKAL OBSERVATIONS AND CONCLUSIONS. 449 



our cereals and to rice to make it worth the trouble of 

 planting it. They had a few bulbs and edible berries, 

 but they have not tried to cultivate them, having early 

 received the maize, which was worth far more. 



Patagonia and the Cape have not furnished a single 

 species. Australia and New Zealand have furnished one 

 tree. Eucalyptus globulus, and a vegetable, not very 

 nutritious, the Tetragonia. Their floras were entirely 

 wanting in graminse similar to the cereals, in leguminous 

 plants with edible seeds, in Cruciferae with fleshy roots.^ 

 In the moist tropical region of Australia, rice and 

 Alocasia niacrorhiza have been found wild, or perhaps 

 naturalized, but the greater part of the comitry sufiers 

 too much from drought to allow these species to become 

 widely difihsed. 



In general, the austral regions had very few annuals, 

 and among their restricted number none ofiered evident 

 advantages. Now annual species are the easiest to cul- 

 tivate. They have played a great part in the ancient 

 agriculture of other countries. 



In short, the original distribution of cultivated species 

 was very unequal. It had no proportion with the needs 

 of man or the extent of territory. 



Article II. — Number and Nature of Cultivated Species at 

 Different Epochs. 



The species marked A in the table on pp. 437-446 

 must be regarded as of very ancient cultivation. They 

 are forty-four in number. Some of the species marked 

 B are probably as ancient, though it is impossible to 

 prove it. The five American species marked D are prob- 

 ably cultivated as early as those in the category C, or 

 the most ancient in the category B. 



As might be supposed, the species A are especially 

 plants provided with roots, seeds, and fruits proper for 

 the food of man. Afterwards come a few species having 



^ See the list of the useful plants of Australia by Sir J. Hooker, 

 Flora Tasmania, p. ex. ; and Bentham, Flora Australiensis, vii. p. 156. 



2g 



