Pollen morphology was used to clarify some taxonomic differences in the classification of the Loteae tribe and the genus Lotiis. The analysis of palynological data by numerical techniques indicated a clear taxonomic separation of many higher categories. Subgenera Symiatiitrii, Sinipeferia and Acrriispon were determined to be most closely related, while subgenus Horackia is separate. Old World Lotits was found to be stenopalynous while New World Loiru was eurypalynous, thereby separating the species into two phytogeographical regions. Various taxonomic treatments were examined and Old World Lotus appeared to be more clearly defined than the relationships between the North American taxa. This is supported by the results of the palynological study. Some taxa considered.as Lottts sensu strict0 may require classification into other genera. The results indicated affinities within or between taxa the subgeneric and specific level which should aid in selecting possible taxa combinations for purposes of plant breeding and molecular studies. The Loteae is an important tribe in the Leguminosae because many taxa are agriculturally desirable and ecologically significant particularly in the genera Lotus, Tetragonolobrrs and Dorycnium (Duke 1981). Some taxa are hay; pasture and cover plants and supply excellent forage or browse (Grant & Marten 1985). Many species of this tribe do not cause bloat in cattle (Jones & Lyttleton 1971). Allen & Allen (1981) reported that 55 species of Lotus, 5 species of Tetrngortolobris and 5 species of Dorycrzirrrn showed root nodulations indicative of nitrogen fixing ability through association with the bacteria genus Rhizobiwn.In the absence of a monographic treatment of Lotits which included the North American taxa, it seemed advisable to study the taxonomy-palynology of the Loteae. We therefore, investigated the palynology of those taxa which have been shown by different authors to be related at ?ither the generic or specific level to the tribes Loteae or Zoronilleae. All Lotics and Loteae taxonomies are contro-(ersial because of attempts to include or exclude taxa indig-:nous to North America, or Africa (Greene 1890, Brand 1898. Modern authors (Polhill 1981, Isley 1981 have ended to ignore this problem. It can be observed in Table I hat not only are tribal differences inconsistent, but many jiverse generic interpretations of taxa also exist. In spite of hese differences, the taxonomy at the species level is relaively consistent [e.g. see treatment of Greene (1890), Otley (1923Otley ( , 1934, Wiggins (1980) and Isley (1981)l. Practia1 considerations may well be the cause of a lack of taxoiomic agreement at the generic level. Several reasons may be advanced for this problem: 1) taxonomic tradition, e.g. adherence to names complying with the "Code of Botanical Nomenclature"; 2) the reliance of taxonomists on the Engler & Prantl System (Taubert 1893); 3) the occurrence of endemic Loteae-like taxa at several "centers of origin" or "centers of diversity" has caused researchers to create many categ...