40 
1. Cananga syl vest r is prima rive trifaUata (Malaice Oetan). 
2, Cananga sglvestrts secunda dve angns/ifolia. 
o. Cananga aylcestris ta fia sire lati folia. 
Of the first two Rumphius gives figures on t. G(> of the flame volume ; 
and judging from these figures, the plants fall into the modern gentiB 
Polyalihia. 
Linnauis' Species Plant-arum was published in 1753, therefore 
Rumphius' names are in point of time, as they are in point of form, pre- 
Linuaeau. Linnams tktes not accept Gananga as a genua and he refers 
to the Cananga of Rumphius only in a note under Uvaria Zeylanica. 
And the first botanists to adopt the Cananga of Rumphius as a genus 
are Hook. fil. and Thomson (in PI. Tml. ISO). But in 1775 Aublet (iu 
his Ilistoirt; de* Planfrs de la itniane Fraih'aise,) published, in regular 
Linnrcan fashion, the genus Cauangn for the reception of a single species 
nnmed 0. ourei/ow of which he gave a figure (t. 244). Nineteen years 
later (1794) Ruiz and Pavon, (in their Prodromus Florm Peruvians 
et Chilensis,) published under the name of Guatteria a genus with 
exactly the same characters as Aablet's Cananga. Unless therefore 
Hook f. and Thomson are right in making a special case in establishing-, 
as a genus in the Linnnenu sense, the Cananga of Rumphius, Aublet's 
genus Cananga must stand, and to it must be relegated all the American 
species referred to Ruiz and Pavon'a genus Guatteria. Authorities vary 
in their treatment of the Cananga of Rumphius. Dunal (in his Mono- 
graphic- de la famille de* Anonacees) pronounces for the suppression of 
Aublet's Cananga in favour of that of Rumphius who, he incorrectly 
says, assigned two species to it; the fact being as already shown, that 
Rumphius divided Cananga into (a) cultivated (with one sort) and 
(i») wild (sylvestres) with three sorts. Dunal (and I think wrongly) 
refers all the Cananga of Rumphius to Unona. In their Genera Planta- 
ruin, Mr. Benthara and Sir J. D. Hooker retain the Cananga of Rum- 
phius and reduce Cananga of Aublet to Guatteria, Baillon, on the other 
hand, retains the Cananga of Aublet as a genus, and to it. refers all tho 
S. American species of Guatteria. He reduces Cananga odoraia H. f. 
and Th. to Unona and, altering the termination of its generic name, he 
makes it a section of Unona under the sectional title of Cananginm. 
The grounds for separating Cananga from Unonana a genus are thus 
stated by the authors of the Flora Tndica. " In habit and general ap pea r- 
anco this genus closely resembles Unona j but the indefinite ovules pre- 
vent its being referred to that genus. Tho peculiar stamen (with a 
long conical apical point) and tho seeds are themselves, we think, suffi- 
cient to justify us in distinguishing it as a genus." The simplest 
solution of the synonymic knot, and one fur which there is some justi- 
289 
