|Abstract||In the present revision the tribe Millettieae comprises 43 genera, 4 of which are either new, or former subgenera or sections here raised to generic rank: Austrosteenisia, Endosamara, Imbralyx, and Paraderris. The genera Brachypterum, Callerya, Deguelia, Philenoptera, and Sarcodum, generally not accepted for 50 years or longer, have been reinstated. The genera Derris, Lonchocarpus, and Millettia received a more restricted circumscription than usual. Ostryocarpus includes also the commonly accepted genera Aganope and Xeroderris. The rather common and rather widespread genera Muellera and Pongamia have been reduced to Lonchocarpus and Millettia respectively.|
New combinations on specific rank are limited to the type-species of new or hitherto untypified genera, viz. Austrosteenisia blackii, Callerya nitida, Endosamara racemosa, Imbralyx albiflorus, Paraderris cuneifolia. One more combination, Afgekia filipes, not a type-species, is made.
All genera as here conceived are defined by a monothetic set of characters, most genera have also unique characters. The genera Apurimacia, Craspedolobium, Derris, Lonchocarpus, Margaritolobium, Millettia, Paraderris, Philenoptera, Platycyamus, and Requienia lack unique characters.
A key is presented, also containing genera that are likely to be mistaken for members of the Millettieae. The genera are presented in alphabetical order as no unambiguous subdivision of the tribe could be made.
Two different (but complementary) numero-cladistic methods have been applied (chapters 5, 6, and 7). Zandee’s three-taxon-statements-permutation method resulted in numerous possible cladograms with many parallels. With a more restricted selection of characters it did not result into a fully resolved cladogram. Meacham’s compatibility analysis, slightly modified, resulted in three different meagre cliques of mutually compatible characters, therewith exactly indicating the cause of the complexity already intuitively recognized.
The tribe is paraphyletic and, in order to obtain a more satisfactory picture of its natural structure (if attainable at all in this tribe), similar treatments will have to be made of the “surrounding” tribes Dalbergieae, Bossiaeeae, Brongniartieae, Robinieae, Phaseoleae (at least its subtribe Glycininae s.l.), and Abreae. Also more (new) characters have to be found and considered, as the characters used in this study have insufficient mutual compatibilities.