218 du, m. t. masters on south-a.ret0ak iiestiace.e. 



Htpolje:n^a. 



As originally constituted by Brown (Prodr. p. 251) this genus 

 comprised two Australian species, to which Nees (in Linna?a, 

 V. p. 663) added a third from the Cape. Amongst its allies this 

 genus is distinguished by the male plants, which are like those of 

 BestiOj while in the solitary spicate female flowers it is more 

 nearly allied to Ilypodisciis and Willdenovia^ from both of which 

 it is separated by its sessile fruit, destitute of disk-like appen- 

 dages either above or below. From the last-named genus it 

 differs especially in the inflorescence. 



Under this head I have placed a few male plants, for the 

 most part collected by Burchell, and which I have not hitherto been 

 able to match. In most of the species the male spikelets are 

 placed with their edges directed towards the rachis or main axis, 

 the florets themselves, in tliat case, having their backs turned 

 against the axis ; but, in a few cases tlae spikelets are placed 

 back to the rachis, as in Triticitm, the florets then being edge- 

 wise. "When this last arrangement exists, it usually happens that 

 the outermost lateral glume is larger than the rest, having more 

 room to expand. "Where the spikelets are placed with their 

 edges to the axis, and the flowers have their backs placed against 

 or parallel with, it, the form of the perianth, and of the outer 

 glumes especially, seems to be modified according to the degree 

 of pressure exerted on them, or the amount of space left 

 for their development. If the bracts are tightly packed, and the 

 outer glumes are large, as in IT. impolita, then the whole flower 

 is concave on the dorsal aspect, so as to fit on to the axis, and 

 the two outer lateral glumes are by this curvature of the flower 

 thrown back against the rachis, the anterior glume of the outer 

 series (which is smaller) being thrust forwards. In other in- 

 stances, where the perianth is relatively smaller, there is not so 

 much impediment offered to its growth, and the glumes in that 

 case are nearly equal in size, and the flower is not concave 

 posteriorly. 



"WiLLDEKOVIA. 



Taken as originally proposed by Thxmberg (Act. Holm. 1790, 

 p. 28) this genus is natural and well defined. Later writers, such 

 as Endlicher and Nees von Esenbeck, impaired its character by 

 including within it plants that have only a slight aflinity to it. 

 Kunth, however (En. iii. p. 453), takes the genus in the Thun- 

 bergian sense, though his description is not sufficiently accurate. 



