The problem of determining affinity among glossopterid gymnosperms is beset by deficiencies in preservation, natural dissociation of parts, and scarcity of features assuredly critical for morphologic comprarison. The glossopterids probably are not a very heterogeneous group of plants, but this is difficult to prove. The Gondwana glacial "hiatus" has resulted in the omission of a critical chapter glossopterid evolution. As a consequence, morphologic features and phyletic probabilities must be evaluated on a much more hypothetical basis than would otherwise be justified. Confusion has arisen from the lack of morphologic terms that permit clear discussion of a newly evolved type of reproductive structure in glossopterids. The structure, here designated a "fertiliger", consists of a leafy bract, a partially adnate stalk, and a fertile head or capitulum. Seven types of fertile structures are discussed, all of which are bilaterally symmetrical and have different features on dorsiventral surfaces. I regard all fertiligers as ovulate but this interpretation may bot be acceptable to some workers; others may not accept dorsiventral organization of the capitulum as being fundamental. Among glossopterids, however, in spite of differences in preservation that may seem to support a variant interpretation, these ovulate fertiligers are the distinctive features that show general consistency. A single fertile bract bearing several capitula, as exemplified by Lidgettonia, is called a compound fertiliger. Staminate structures (microsporophylls) of glossopterids are separately classified as Eretmonia, Glossotheca, and possibly as other taxa. Only the manner of sporangial attachment is not entirely clear. It seems likely the staminate parts have previously been confused with scale leaves and are actually coextensive in distribution with the glossopterids. A tentative phyletic model suggests the distant derivation of glossopterids from middle Carboniferous cordaiteans. Many details must be speculative due to the lack of a pertinent fossil record, but this interpretation accounts for some features that have no counterpart in pteridosperms. Permineralized ovules from Antarctica provide general support for this working hypothesis, but specific evidence is lacking. Furthermore, it seems unlikely angiosperms originated from glossopterids; it is more reasonable to consider the glossopterids as possible distant ancestors of the Gnetales. ?? 1976.