42 HISTORY AND DISTRIBUTION OF SORGHUM. 



saccharatus is described as " glumis glabris seminibus muticis," based 

 on the Milium, indicum grants flavescentibus of Hermann and the 

 Milium indicum sacchariferum altissimum seminibus ferrugineo of 

 Breyne. In the second edition, 1763, it is described as having an 

 erect, snbverticillate panicle with wide-spreading or somewhat droop- 

 ing branches, in contrast with the small, erect, and ovate heads of 

 Holcus sorghum. Both species are credited with villous glumes and 

 awned lemmas (flowering glumes), and are distinguished by the size 

 and character of the panicle. 



In his " Mantissa " (1771) he restricts the name Holcus sorghum to 

 a strain of the ovate-panicled form with green villous glumes, and 

 adds Bauhin's form with flat white seeds as a synonym. He then per- 

 mits the open-panicled Holcus saccharatus to have hairy glumes and 

 awned lemmas also, and describes a new species, Holcus bicolor, simi- 

 lar to H. sorghum, with glabrous black glumes and globose white 

 seeds, in awned lemmas, transferring his Holcus glumis glabris of 

 1737 to Holcus bicolor as a synonym. This new species was said to 

 have come from Persia, which probably indicates that it was an In- 

 dian variety perhaps secured through Persia. A similar form is 

 found abundantlv in India to-dav- It could scarcelv have been the 

 white dnrra of Arabia, etc., because that had pale glumes. 



Forskal (1775) added two species from Egypt: Holcus dochna 

 and H. durra, the specific names being derived from the common 

 names of the varieties in that country. Holcus dochna was probably 

 a saccharine or semisaccharine sort, but is not now certainly identi- 

 fiable. Both Koernicke (1885, p. 310) and Hackel (1889, p. 509) 

 place it as a synonym under saccharatus. Forskal does not mention 

 it as saccharine, but the description accords well with that of the 

 sweet forms. His Holcus durra included at least two of the three 

 durra varieties common in Egypt to-day, namely, the white seeded 

 (beda) and the brown seeded (ahmar). Indeed, Forskal makes it 

 cover two white-seeded forms, one with greenish glumes and one with 

 brownish glumes. Koernicke (1885, p. 312) places the reddish -brown- 

 seeded sort under the variety arduini Gmel. In 1887 he proposed the 

 name egyptiacus for the white-seeded form. Hackel follows him in 

 this, placing under his variety durra only a yellow-seeded (safra) 

 form; he does not seem to have had the brown-seeded form at all. 

 These three forms, beda, safra, and ahmar, with white, yellow, and 

 brown seeds, respectively, are in common cultivation in Egypt to-day. 

 They are practically identical except for the color of the seed, and 

 really represent but a single botanical variety instead of three. 



The Species of Arduino. 



Arduino (1786) published descriptions of six species of sorghum, 

 three of which he considered new, the other three being those de- 

 175 



