PROC. BIOL. SOC. WASH. 92(4), 1979, pp. 689-696 COMMENTS ON THE PHYLOGENY OF PERCHING BIRDS Alan Feduccia Abstract. — The bony stapes (columella) is a unique character in birds in that the primitive condition is the retained reptihan condition of the element, and derived ''pockets" of stapedial morphologies occur that may be of importance in clarifying certain phylogenetic relationships. Scanning elec-tron micrographs (SEM's) of various views of stapes of suboscines, certain coraciiforms and trogons are here presented and reveal with clarity the manifest differences betweeen suboscines and the alcediniform coraci-iforms. These new data, combined with new data from sperm morphology of suboscines, suggests that the Passeriformes is a monophyletic assem-blage, that the suboscine and alcediniform stapes evolved independently, and that the suboscines are more closely related to the oscines than to the alcediniforms. In a previous paper (Feduccia, 1977), I proposed a hypothetical phylogeny of the passerines and related avian groups, birds that have presented difficult phylogenetic problems because of their morphological uniformity and the probability of massive convergent evolution. My reconstruction was based primarily on the discovery of a new avian taxonomic character, the config-uration of the bony stapes or columella, which had until recently been over-looked because of its minute size (one to several mm) and its location in the recesses of the middle ear cavity. In addition, it is often broken or lost in skeletal preparations. The avian stapes is perhaps unique as an avian morphological character in that the primitive condition is the retained rep-tilian stapedial morphology, which is characteristic of most groups of birds. In some groups, however, the stapes exhibits pecuHar derived morpholo-gies. Where these ''pockets" of derived stapedial morphologies are found they are considered as strong indications of evolutionary affinity, unless there are compelling reasons to assume that the morphological similarities are due to convergent evolution. I made the following broad conclusions (Feduccia, 1977). First, the hoo-poes (Upupidae) and wood-hoopoes (Phoeniculidae) are monophyletic with-in the Coraciiformes {sensu Wetmore 1960); previously there was consid-erable dispute as to their relatedness. These two families are characterized by a stapedial morphology that is termed the "anvil" stapes and is illustrated in Fig. 1. Second, the coraciiform families containing the bee-eaters (Mer-
BioStor
