364 MK. O. BENTHAM ON CYPEBACE.E. 



of what constitutes a species or a genus. Wherever lie could not 

 readily determine a specimen he had before him, he at once 

 described it as new, thus adding in no small degree to such 

 confusion as Nees and others may have created. 



The last Cyperographer I have to mention is Boeckeler, who 

 seems to have a thorough knowledge of species and (in several 

 volumes of the ' Linnsea ' and of the Ratisbon ' Flora ') has elabo- 

 rately described all those of the Berlin herbarium and of various 

 collections submitted to him for the purpose. In so doing he has 

 verified a number of synonyms, and often successfully rearranged 

 some of the species of the larger genera. His labours, however, 

 would have been much more useful if he had had access to more 

 of the works where Cyperacese had been described, if he had not 

 indulged so much in page-long uncontrasted diagnoses, if he had 

 had rather a clearer idea of the principles upon which generic and 

 ordinal distinctions are founded, and if he had adopted a more 

 correct terminology. In several instances he has quite mistaken 

 Brown's genera. The new ones he has at various times proposed 

 have mostly proved, even upon his own admission, to be common 

 species of well-known genera ; and amongst those he has retained, 

 one (Lasiolepis) is founded upon three not uncommon species 

 of Eriocaulon. As to terminology, he has, for instance (chiefly 

 after the example of Nees), given the name of perigynium to a 

 glume in Carex (Linnsea, xxxix. 1), to that which is supposed to 

 replace the perianth outside of the stamens in Lepidospermum, to 

 the gynophore or hypogynous disk inside of the stamens in Ficinia 

 and Scleria, and to the pericarp in Anosporum, a supposed genus 

 compounded of three Cyperi and a Scirpus. 



Boeckeler's primary division of the order into two series, 

 according as the fertile flower is hermaphrodite or strictly female, 

 bears the test of a detailed examination ; for in those species of 

 Caustis and a few other Bhynchosporece which are said to be 

 dioecious, the fertile flow r ers are female only by the abortion of 

 the anthers, the filaments are present as staminodia, at least in 

 all the flowers I have examined, whilst in the strictly unisexual 

 genera there is no trace of staminodia in the females or of a rudi- 

 mentary ovary in the males. Boeckeler has, however, made the 

 mistake of including in Carpha the strictly unisexual genus 

 Eriophora, Hochst. 



The three tribes into which the hermaphrodite series of genera 

 are distributed (Kunth's four tribes reduced to three by the union 



