TERMINOLOGY IN MONOCOTYLEDONS. 505 



each a single flower. If, as stated by Hieronymus, in his careful 

 and elaborate monograph of the Order, these flowers were really 

 all constantly solitary and hermaphrodite in A. cyperoides, that 

 would be an example of a typical spikelet with distichous glumes ; 

 but in the specimens I have examined I have always found two 

 flowers, one or both male only, in the lowest bract, and sometimes 

 also in the second; and in all the other species artificially sepa- 

 rated by Hieronymus under tbe name ofBrizula, the lowest one or 

 two bracts contain two or more flowers, usually males, and each 

 of the other bracts a single flower only, but a female one. In 

 Centrolepis there are usually only two primary bracts ; but the 

 flowers are generally two or more together within each, above 

 twenty in C. Banksii and C. exserta, and so densely crowded on so 

 short an axis that their precise arrangement is not easy to trace. 

 Hieronymus describes this partial group as a centrifugal cicinnus 

 (a cyme reduced to a single branch). I have, however, felt con- 

 vinced, especially after a careful examination of those species 

 where it is most developed, that it is, on the contrary, a centri- 

 petal, unilateral, or rather secund spike— that is, that the flowers, 

 though affixed round the short axis, are all turned to one side, the 

 lowest being the first developed. Under each flower, in most of 

 the species, are one, two, or three very thin and slender hyaline 

 scales ; as one of these is almost always immediately under a 

 stamen, and the other one or two more lateral, they may possibly 

 represent a reduced perianth, but are perhaps more likely to be 

 bracts and bracteoles. Upon the whole, however, with this un- 

 certainty as to their homology it may be better in Centrolepideas 

 to make no use of the term glume, but to retain the name of bract 

 for those which are on the primary axis of the inflorescence, and 

 to designate as scales the hyaline ones within the bracts. In the 

 little genus Trithuria the minute flowers are numerous in a 

 nearly globular head, as in Eriocaulon, but with only a few bracts 

 towards the circumference, the stamens and pistils being densely 

 crowded in the rest of the head without intervening bracts or 



scales. 



In the great majority of Kestiaceae the glumal arrangement is 



normal, whilst the perianth, although not protruding beyond the 

 glumes, is as regularly developed with its six dry or scale-like 

 segments as in the Junce* ; and yet, by some singular perversion 

 of terms, these perianth-segments are by several botanists de- 



name 



