INTRODUCTION Despite their fascination for mankind, the venomous snakes remain incompletely known as to numbers of species, relationships within and among families, and even the basic anatomy of the venom-injecting apparatus. This paper is concerned with using anatomical materials to help unravel one of the evolutionary puz-zles in the viperid snakes. It is an outgrowth of a taxonomic study of the viperine snakes by Marx and Rabb (1965), who commented on relationships of three odd, apparently primitive genera: Causus, Azemiops, and Atractaspis. Except for Causus, the anatomy of these forms was little known at the time. Bourgeois (1965) has since established that Atractaspis, although possessing a folding-fang venom-injection apparatus, is not a viper but rather has aparallactine colubrid affinities, a viewpoint further substantiated by the work of Kochva et al. (1967) and McDowell (1968). We have subse-quently concentrated attention on Azemiops, a rare monotypic genus from temperate montane areas of southeastern Asia (Boulen-ger, 1888; Pope, 1935; Bourret, 1936). The primitive nature of Azemiops is reaffirmed in a concurrent study of various skeletal and integumental characters in taxa representing all the advanced snake groups (Marx and Rabb, in press). To elucidate the position of Azemiops among the Viperidae, we have examined in detail parts of the head musculature and certain skull elements. The descriptions form the bulk of this paper. The study includes functional interpretations of the mor-phology, which we believe contribute to a fuller understanding of the mechanics and evolution of the viperid skull. Analysis of our findings as a whole indicates a distinctive intermediate phylogenetic position for Azemiops. We propose that this position be recognized by establishing a separate subfamily to contain this singular genus. 67
BioStor
