106 THE JOURNAL OF BOTANY 



tliroLighout the Linnean literature. Lamarck (Encycl. Metli. ii. 147 

 (1786), gives a condensed account, based entirely upon Aublet's 

 descriptions; both the Gaertners (DC. Fruct. 1788-1805) leave 

 Coupoui without mention, despite the careful description of the fruit 

 available; so also A. L. de Jussieu (Gren. PL 1789). In the Pro- 

 dromus (iii. 295, 1828) De Candolle includes Coupoui among his 

 " Myrtaca? dubise," with the remark " Flores ignoti." Bentham and 

 Hooker were the first systematists to guess the famil}^ to Avhich this 

 genus seems rightly to belong — Huhlaceoe ; and in the absence of 

 flowers their determination amounts to little more than a guess — 

 as Avitness their erroneous suggestions of its affinities Avithin the 

 family in question. Thus Bentham {Qen. Plant, i. 696, 1865) 

 removes Coupoui from Myrtacese and relegates it to Pentafpnia, in 

 the group GoteshceecG of Rubiacece — a tribal afiinity inadmissible 

 on the ground of fruit-characters alone, apart from the sex- separation 

 and the sestiv^ation of the corolla. 



Miers's treatment of Aublet's nomenclature is more than drastic 

 and, to modern ideas, is unpardonable. He alters both names, generic 

 and specific, advancing reasons that seem altogether inadequate {op. 

 cit. 16). Coupoui, he fears, might be confounded with Coupeia, 

 Ooupia, and Cupia, so he proceeds to create real confusion by naming 

 the genus Cupirana, i. e. " wild " Cupi, " to distinguish it from the 

 true Cupi ( Gouepia guianensis) " — a distinction of which the necessity 

 was unsuspected previously. "Aublet's specific name," continues 

 Miers, *' is not adopted, as it presupposes {sic) a small aquatic plant." 

 Baillon, in his Hisioire des Plantes, x. 176 (1888), restores the 

 name Coupoui, and includes the genus among doubtful Apocynacece, 

 following Miers's account without making any personal research or 

 comment. 



In the matter of nomenclature Eafinesque, in his Principes 

 fondamentaux de Somiologie, etc. (1814), was more merciful than 

 Miers. On page 29 of this interesting work he enunciates a " Begle. — 

 Quand la terminaison des noms geneiiques est barbare, il faut la 



modifier Ohs. — Ainsi il faut ecrire . . . Cupuia au lieu de 



Coupoui. . ." 



In all the circumstances it would seem best to follow Baillon in 

 adhering to the original Aubletian name, leaving any change for a 

 future Botanical Congress ; Coupoui appears to be no more "barbaric" 

 than man}^ other names that still stand without question. 



I have examined the flowers of Martin's specimens referred to 

 above ; and, apart from their superficial appearance and contorted 

 corolla-lobes, there is little reason to associate them \w\t\\ApocynacecB. 

 Above all, the sexes are separated — a rare, if not unrecorded, condition 

 in the last-named family. I have found no female flowers, unfor- 

 tunately, upon any of the Coupnui specimens that I have seen ; all 

 are staminate, with sterile female parts at most. Moreover, all these 

 male flowers are hexamerous ; Miers, in his generic description, wrongly 



describes them as pentamerous — " corollas limbi segmenta 5 



stamina 5 " In fact, hexamery appears to be a character 



constant for the male flowers of all the species ; while from Aublet's 

 description of the fruit, it would seem that the female flowers are 



