382 ORIGIN OF CULTIVATED PLANTS. 



argues an introduction of but few eentixries before the 

 Christian era. 



No botanist mentions the dourra as wild in Egypt 

 or in Arabia. An analogous form is wild in equatorial 

 Africa, but R. Brown has not been able to identify it,^ 

 and the flora of tropical Africa in course of publication at 

 Kew has not yet reached the order Graminse. There 

 remains, therefore, the single assertion of Dr. Bretsch- 

 neider, that the tall sorghum is indigenous in China. 

 If it is really the species in question, it spread westward 

 very late. But it was known to the ancient Egyptians, 

 and how could they have received it from China while 

 it remained unknown to the intermediate peoples ? It 

 is easier to understand that it is indigenous in tropical 

 Africa, and was introduced into Egypt in prehistoric 

 time, afterwards into India, and finally into China, where 

 its cultivation does not seem to be very ancient, for the 

 first work which mentions it belongs to the fourth cen- 

 tury of our era. 



In support of the theory of African origin, I may quote 

 the observation of Schmidt,^ that the species abounds in 

 the island of San Antonio, in the Cape Verde group, in 

 rocky places. He believes it to be " completely natural- 

 ized," which perhaps conceals a true origin. 



Sweet Sorghum — Holcus saccharatvs, Linnteus ; An- 

 dropogon saccharatus, Roxburgh ; Sorghum sacchara- 

 tum, Persoon. 



This species, taller than the- common sorghum and 

 with a loose panicle,^ is cultivated in tropical countries 

 for the seed — which, however, is not so good as that of 

 the common sorghum — and in less hot countries as fodder, 

 or even for the sugar which the stem contains in con- 

 siderable quantities. The Chinese extract a spirit from 

 it, but not sugar. 



The opinion of botanists and of the public in general 

 is that it comes from India ; but Roxburgh says that it 

 is only cultivated in that country. It is the same in 



' Brown, Boi. of Congo, p. 544. 



' Schmidt, Beitrage zur Flora Gapverdischen Inseln, p. 158. 



• See Host, Oraminai AustriacoB, vol. iv. pi. 4. 



