Both the anguli and externus superficialis of snakes, as labelled by previous authors, are together the homolog of the lizard levator anguli oris; the externus profundus of snakes, as identified previously, is the homolog of the lizard adductor externus superficialis; the true adductor externus profundus, which would run deep to the facial artery but superficial to the trigeminal nerve, is absent in snakes. The rictal fold of lizards is much shortened in snakes, at least at its aperture, and becomes the tunic of the venom gland in protero- glyphs (including Atractaspis), viperids, and Gonionotophis and Mehelya (Colubridae). The spe- cialization in secretory function of some oral epithelium is probably an independent event, devel- opmentally, from the organization of the gross form of the venom gland by mesenchyme and a muscularized venom gland is the result of the specialized epithelium entering the rictal fold as a lining; if the specialized epithelium remains external to the rictal fold, it forms a Duvernoy's gland. A series of intermediates, currently classified in Calliophis, connects typical elapids with Atractas- pis and the Homoroselaps Group (essentially, the Aparallactinae but without Aparallactus and Macrelaps). The musculature of the venom gland of typical elapids appears to be the levator anguli oris, attached to the rictal plate; this attachment seems a primitive, rather than specialized, feature and is suggestive of uropeltids that are considered primitive snakes, in large part because of this muscle pattern. The different patterns of venom gland musculature among typical proteroglyphs can be explained as the interaction of two quantitative developmental factors: posterior extent of the levator anguli oris; and dorsal growth of the levator anguli oris relative to the adductor externus medialis. Proteroglyphs are more primitive than colubrids in their head musculature and in failure to bring the fangs close to the attachment of the pterygoideus superficialis. Some proteroglyphs have extended non-venomous maxillary teeth close to the pterygoideus attachment on the ecto- pterygoid and provide some insight into how this may have happened in colubrids. The colubrid rear-fanged condition involved backward extension of the maxillary tooth row, with an accompa- nying Duvernoy's gland formed from the dental lamina, on the homolog of the hind-maxilla of Bolyeridae, a portion of the maxilla characterized by special ligamentous connections to the palatine and not present in booids, anilioids, acrochordoids, or tropidophids (present, but not fang-bearing, in proteroglyphs). The viperid condition seems derived from the colubrid, but the characteristic posterior medial process of the maxilla and its ligament to the lateral proces of the palatine have been lost and the serous secretory epithelium has entered a secondarily enlarged rictal fold (venom gland) that has acquired a new musculature not utilizing the levator anguli oris. The primitive nature of proteroglyphs, as compared to colubrids, suggests that the most primitive colubrids should be looked for among the most proteroglyph-like forms and Homalopsinae appear to fit this crite- rion.