Evolution

The phylogeny of snakes is poorly known because snake skeletons are typically small and fragile, making fossilization unlikely. It has however been generally agreed, on the basis of morphology, that snakes descended from lizard-like ancestors. Recent research based on genetics and biochemistry confirms this; snakes form a venom clade with several extant lizard families.

Recent fossil evidence suggests that snakes directly evolved from burrowing lizards, either varanids or some other group. An early fossil snake, Najash rionegrina, was a two-legged burrowing animal with a sacrum, fully terrestrial. One extant analog of these putative ancestors is the earless monitor Lanthanotus of Borneo, although it also is semi-aquatic. As these ancestors became more subterranean, they lost their limbs and became more streamlined for burrowing. Features such as the transparent, fused eyelids (brille) and loss of external ears, according to this hypothesis, evolved to combat subterranean conditions (scratched corneas, dirt in the ears). According to this hypothesis, snakes re-emerged onto the surface of the land much as they are today. Other primitive snakes are known to have possessed hindlimbs but lacked a direct connection of the pelvic bones to the vertebrae, including Haasiophis, Pachyrhachis and Eupodophis) which are slightly older than Najash.

Primitive groups among the modern snakes, pythons and boas, do have vestigial hind limbs, tiny, clawed digits known as anal spurs and used to grasp during mating. Leptotyphlopidae and Typhlopidae are other examples where remnants of the pelvic girdle are still present, in Leptotyphlopidae sometimes as horny projections or not visible at all. The frontal limbs in all snakes are gone because of the evolution of the Hox genes in this area. The axial skeleton of the snakes' common ancestor had like most other tetrapods the familiar regional specializations consisting of cervical (neck), thoracic (chest), lumbar (lower back), sacral (pelvic) and caudal (tail) vertebrae. But the Hox gene expression in the axial skeleton responsible for the development of the thorax became dominant early in snake evolution. As a result, the vertebrae anterior to the hindlimb buds (when present) all have the same thoracic-like identity (except from the atlas, axis and 1-3 neck vertebrae), meaning most of the snake's skeleton is actually made up of an extremely extended thorax. Ribs are found exclusively on the thoracic vertebrae. The neck, lumbar and pelvic vertebrae are very reduced in number (only 2-10 lumbar and pelvic vertebrae are still present), while only a short tail remains of the caudal vertebrae, although the tail is still long enough to be of good use in many species, and is modified in some aquatic and tree dwelling species. Because the front (thoracic) limbs in tetrapods appear in the area between the neck and the thorax, a location that is now almost absent in snakes, there is simply no longer any room left where they can develop.

The alternative hypothesis, based on morphology, suggests that ancestors were related to mosasaurs — extinct aquatic reptiles from the Cretaceous — which in turn are thought to have derived from varanid lizards. Under this hypothesis, the fused, transparent eyelids of snakes are thought to have evolved to combat marine conditions (corneal water loss through osmosis), while the external ears were lost through disuse in an aquatic environment, ultimately leading to an animal similar in appearance to sea snakes of today. In the Late Cretaceous, snakes re-colonized the land much like they are today. Fossil snake remains are known from early Late Cretaceous marine sediments, which is consistent with this hypothesis, particularly as they are older than the terrestrial Najash rionegrina. Similar skull structure; reduced/absent limbs; and other anatomical features found in both mosasaurs and snakes lead to a positive cladistical correlation, though some features are also shared with varanids. Supposedly similar locomotion for both groups is also used as support for this hypothesis. Genetic studies have indicated that snakes are not especially related to monitor lizards, and (it has been claimed) therefore not to mosasaurs, the proposed ancestor in the aquatic scenario of their evolution. However, there is more evidence linking mosasaurs to snakes than to varanids. Fragmentary remains that have been found from the Jurassic and Early Cretaceous indicate deeper fossil records for these groups, which may eventually refute either hypothesis.

The great diversity of modern snakes appeared in the Paleocene, probably correlated with the adaptive radiation of mammals following the extinction of the dinosaurs.