70 



(as to congesta) " bifid, shortly cuspidate between lobes tend- 

 ing to the shape of a knife, having the appearance of a triple 

 crown." The upshot of this is merely that the barren sta- 

 mens in one are entire, and in the other split into two acute 

 lobes at the end, with sometimes an irregular side-tooth, a 

 very good specific feature. Who ever thought of separating 

 Iris into genera on account of the indentures of the margin of 

 its crests ? I shall perhaps surprise the reader by stating, 

 that, as far as I know, the asserted hypogynous scales in 

 Brodisea seem to be a fallacy, and that no such thing has 

 existed in any of the flowers I have examined of either species. 

 The perianth is thick, and Salisbury was deceived by remain- 

 ing fragments thereof when he thought he had pulled it ofi\ 

 and others have, I suppose, taken them for granted. There- 

 fore "squamse hypogyna? nullas" in other genera is superfluity. 

 The supposed scales in Pyrolirion, which deceived Ruiz, were 

 an articulate base to the alternate filaments. The difference 

 stated by Kunth that B. congesta has the sepals, and grandi- 

 flora the petals, widest, is incorrect : the petals are widest in 

 both. Prof. Endlicher places Triteleia and Hesperoscordura, 

 with genera intervening between them, in Agapanthea?, and 

 Allium in Asphodelese, dividing the original Asphodelese into 

 suborders, which Prof, Kunth wisely, (because they are not 

 correct) but, I believe, silently, rejects. On examination of 

 the characters of those new suborders, (as well as of Aloinea? 

 which intervenes) it will be found that there is no true dis- 

 tinction, the one suborder by alternatives admitted in its 

 character comprising the points to which the other is limited, 

 while other distinctions are incorrectly assumed. The only 

 positive difference asserted is in the seeds, and that in some 

 respects inaccurately, in others insignificantly. It is not a 

 fact that his Agapantheee have a black or pale membrana- 

 ceous, and his Asphodelese a black crustaceous shell, nor are 

 those distinctions true even as to genera. Some of the Cape 

 Ornithogala have a less crustaceous shell, more compressed 

 and less globose, than most other Ornithogala, or the genera 

 Triteleia, Brodisea, and above all Calliprora, in his Agapan- 

 thcDS ; and Scilla amccna in his Asphodelese has the shell 

 rufous brown amongst its black-seeded congeners. 



W. Herbert. 



