ii2 



Moreover tlie Lauracese^^ of tlie Ranales sliow a close affinity to tlie Tliy- 

 melsBacssB in the thoroughly circular aiTangement of the floral elements in tha 

 psrigynous insertion of the stamens, and in the symcarpous gyngeceiim. 



Next, come the Rhoeadales. Of this series, Engler speaks as foUows"^^ : — 



Schon lange hat man die Rh.oeadales gern an die llnmles angeschlossen, un:l 

 (lies hat auch seine Bereclitigung, da die Papavernceo'. wegen ibrer me-st zahlreichen 

 Staubbliitter und ihres bisweilen noch aus mehreren, wenn auch vereinten Carpellen 

 gebildeten Gyniiceums Analogien mit den NymplumcerK zeigen, bei welchen ein syn- 



carpes Gyniiceum zu Stande kommt frehr groBe Ubereinstimniung zeigen 



mit dieser Keihe die 1 'arietales, von denen namenthch die LVMn^acea; an die Jianahs 

 anklingen, wiihrend die Flaconrliacca; zu den Capparidacea; reziehungen ergcben, 



But, in my opinion, the real relation of the Rhoeadales to other seiies extends 

 far more widely. Its affinity to the Centrospermse and to the Ranales has 

 been discussed. Further, the Rhoeadales bear some resemblanca to the 

 Greraniales, the Hkeness being showu in the KoeberHniacese"^^ and Kutacese in 

 the presence of oil-glands and in the other anatomical characters (Benth.- 

 Hooker). The same is true of the former family and the Siraarubacese (Benth.- 

 Hooker). On the other hand, a very close and rather intricate relation is 

 to be found between some famihes of the Rhoeadales and those of the 

 Parietales, namely : — between the Papaveracese and the Loasacese'*^ (De Can- 

 dolle) ; between the Capparidacese and the FlacourtiaceBe^^ or the Erythro- 

 spermeae of the hitter family ; between the MoringaceEe'"'^ and the Violaceae 

 (Hooker) ; between the KoeberHniaceae''^ and the Bixacese in their deciduous 

 sepals ; and between the former family and some others of the Parietales in 

 the placenta with numerous poly-seriate seeds (Benth.-Hooker). Moreover, 

 some connection between the Rhoeadales and the Sarraceniales is indicated in 

 the Papaverac38B and SaiTaceniacese^^. Eurthermore, the agreement of the 

 KoeberHniacese''^ and Saxifragaoese (in the placenta with numerous poly-seriate 

 seeds), and that of the Moringacese^"' and Leguminosse show some relation 

 between the Rhoeadales and the Rosales (Bentham). And finaUy, this 



1) Nat. Pfl.-fam. III.— 2, p. 111. 2) Englee, A. — Erlauterungen, 1. c. p. 365. 



3) Nat. Pfl.-fam. in.— 6, p. 321. 4) Nat. Pfl.-fam. m.— 6, a, p. 106. 



5) 1. c. in.-6, a, p. 10. 6) 1. c. nL-2, p. 243. 



7) 1. c. in.— 6, p. 321. 8) 1. c. Itl.— 2, p. 251. 



9) Nat. Pfl.-fam. III.— 6, p. 321. 10) 1. c. III.— 2, p. 243. 



