MB. G. BENTHAM OS BIXACE.E AIH> SAMTDACBJE. 75 



Notes on BixacecB and Samydacece. By Geobge Ben^ham Esq 



V.P.L.S. 



[Read March 7, 1861.] 

 There is a long series of polypetalous Orders with parietal pla- 

 centation, so closely connected with each other, that systematists 

 have in general endeavoured to keep them together in the linear 

 series, although the doing so interferes much with any definite 

 subdivision of the Class. There is no doubt that the hypogynoua 

 Capparideae, Cistaeeae, Violaceae, and Bixaeea? can only be sepa- 

 rated from the perigynous Samydeae, Homalineaa and Passifloreie 

 by an arbitrary line ; and that this series and that of the curvem- 

 bryonous Orders (Caryophylleae and its aUies) furnish the strongest 

 arguments against the CandoUian arrangement. Yet even in these 

 Orders, the chain which connects them is nearly as closely linked 

 on to others which, under any arrangement, are kept wide apart. 

 Thus Capparideae are almost blended with Crucifera>, Bixaceaa and 

 Samydaceae with Tiliaceae, Violaceae through Sauvagesieaa with 

 Frankeniaceae and Hypericineae, Passifloreaa with Cucurbitacea?, 

 Loasaceae, Turneraceae, &c. Parietal placentation, moreover, can- 

 not be taken as a character of so high an order as it might appear 

 at first sight; for it occurs exceptionally in almost every large 

 calyciflorous Order, as well as in a few Thalamiflor®, which have 

 normally axile placentas ; and it is not constant in several of the 

 above-mentioned Orders where it prevails. 



In our ' Genera Plantarum,' Dr. Hooker and myself have deter- 

 mined, for reasons which we hope to give in detail, to maintain 

 the CandoUian series in its general features; and therefore, in 

 hunting the Orders now under consideration, we have to rely, in 

 the first instance, on staminal insertion. Among those botanists 

 who generally follow these views, Capparideae, Cistineae and Vio- 

 laceae are universally retained under Thalamiflorse, and Passiflo- 

 reae under Calyciflorae ; the intermediate genera have been very 

 variously grouped. DeCandolle distributed them into four Orders, 

 — Bixaceae and Placourtiaceae placed next to each other, in Thala- 

 miflorae ; Samydeae and Homalineae, rather widely separated, under 

 Calyciflorae — but at that time he had had opportunity of examin- 

 ing only a very small number of genera. Endlicher, in his ' Genera 

 Plantarum,' and Bennett, in the ' Plantae Javanicse Bariores,' 

 proposed the union of the two first Orders. Clos, in the ' An- 

 nates des Sciences JNaturelles,' ser. 4. vol. viii., in revising the 

 Order thus combined, unites with it a small group proposed by 



