46 HISTORY AND DISTRIBUTION OF SORGHUM. 



The first section includes five varieties : Cafer Ard., technicus Kcke., 

 saccharatus L., leucospermus Kcke., and niger Ard. The first two 

 varieties are separated from the others by having a much shortened 

 rhachis, or central axis, in the panicle. The variety cafer is our 

 Planter sorgo, or a very similar variety, while the variety technicus 

 is broom corn. The last three are apparently all sorgos or sweet 

 sorghums. The variety saccharatus is much like our most spreading 

 Eed Amber, the variety leucospermus is not determinable, and the 

 variety niger is probably identical with some of our smaller and more 

 compact Amber forms, as Black Dwarf. Koernicke's leucospermus 

 had not white seeds, but red, the inappropriate name apparently re- 

 ferring to the pale glumes, due perhaps to the acknowledged late and 

 incomplete maturity of the specimens on which it was based. 



The second section contains seven varieties, of which the first four — 

 usorum Nees., arduini Gmel., aethiops Kcke., and bicolor L. — had 

 erect heads, and the last three — cernuus Ard., truchmenorum Koch., 

 and neesii Kcke.— had the heads pendent. Of the first group, with 

 erect panicles, the variety usorum, with short compact heads, is not 

 certainly identifiable with any of the hundred or more Natal forms 

 studied by the writer. Variety arduini is probably our Orange sorgo, 

 variety aethiops is not known, though it is perhaps the sorgo form 

 with ovate black heads still grown in central Europe. Variety 

 bicolor is not identifiable among the many similar forms of India 

 and equatorial Africa. Considering, now, the three varieties having 

 pendent heads, it is known that the variety cernuus includes, among 

 others, our white durra, and that the variety truchmenorum is that 

 form of white durra with taller stalks and more compact heads 

 found in Turkestan, where it is called, in Kussian, " Dzhugara." 

 Variety neesii is a blackhulled white durra from the region of Natal. 

 It has never been found in the recent importations from there. 



Hackel (1880) divides Andropogon sorghum into two subspecies, 

 halepeusis and sativus, thus including both the wild and the culti- 

 vated forms in a single species. His subspecies sativus is the Andro- 

 pogon sorghum of Koernicke's classification and the Sorghum vulgare 

 of Persoon. He divides this subspecies into nine sections, which 

 contain a total of thirty-six varieties. These nine sections, or groups, 

 are separated largely on such characters as the comparative shape and 

 size of the spikelets, the comparative length and width of the glumes, 

 and the relation of these to the seed. The density of the panicle, the 

 position of its branches, the color of the seeds, and the presence and 

 length of the awn are used as minor determining characters. No 

 mention is made of the size and height of the culms, the character 

 of the juice, the number or size of the leaves, the comparative length 

 of the sheaths and internodes, or of the purpose for which the 

 varieties are grown. 



175 



