Bell, 1827
Spider tortoises
Recognition
The elongated, hingeless carapace has parallel or almost parallel sides, and is either domed or flattened dorsally. A small notch is present in the cervical region, and the posterior carapace rim is downturned and unserrated (or slightly so in juveniles). The cervical scute may be absent. Vertebrals are broader than long with slightly raised centers surrounded by prominent growth annuli. Neurals alternate in shape between hexagonal or octagonal and square. Eleven marginals are present on each side; the supracaudal is undivided. No submarginal scales are present. The plastron is well-developed and contains a moveable hinge between the humeral and pectoral scutes (between the epiplastra and hyoplastra, either crossing the posterior portion of the entoplastron or behind it) in one species, but not in the other. The forelobe is longer than the hindlobe, which may or may not be posteriorly notched. One or two axillary and inguinal scutes are present on the broad bridge. Axial and inguinal buttresses are short but strong. The gulars are not greatly thickened and extend only slightly beyond the carapacial rim. The head is moderate in size with a nonprojecting snout and a slightly hooked upper jaw. The maxilla may bear a slight median ridge or be ridgeless. The sides of the jaw are serrated in one species but smooth in the other. Limbs are clublike with five claws on each forefoot. Seven to nine longitudinal rows of enlarged, nonoverlapping scales are present on the anterior surfaces of each foreleg. Spurlike scales are present on the heel, and some small blunt tubercles on the thighs. On the anterior surface of each hindleg is a longitudinal row of nonoverlapping scales. The toes are not webbed, and the tail ends in a spinelike scale.
Remark
This genus includes only two small tortoises from Madagascar: Pyxis arachnoides Bell, 1827 and P. planicauda (Grandidier, 1867). Obst (1978, 1980) and Bour (1981a) moved P. planicauda from the genus Acinixys Siebenrock, 1903b to the genus Pyxis, designating Acinixys as a subgenus of Pyxis. This was done on the basis of similarities in carapace color, pattern, and morphology; and skull and shell osteology (see Bour, 1981a, for a complete discussion). There are, however, major differences between the two species in the plastron (hinge versus hingeless, plastral formulae) and the morphology of the cervical vertebrae, and Crumly (1985, pers. comm.) thought many of the of the characters used to support the new alignment those of juveniles shared also with the young of Homopus, Psammobates, Chersina, Kinixys, and even some Testudo and Geochelone. The inclusion of planicauda in the genus Pyxis was also not accepted by Durrell et al. (1989). Obviously, further study is required to determine the true relationship of these two species.