IV. Fresh-water Snakes, distinguished by the position of the nostrils, which are placed on the top of the snout, and by a tapering tail. They inhabit fresh-waters, and are, therefore, excellent swimmers and divers; only a few species (which also in external characters approach the following group, that of the true Sea Snakes) venture out to sea. They feed on fishes, frogs, crustaceans, and other water animals, and are viviparous. Not any of them are venemous.
V. Sea Snakes, distinguished by a strongly compressed tail, and by the position of the nostrils, which are placed as in the last group. They live in the sea, only occasionally approaching the land, feed on marine fishes, are viviparous and venemous. One genus only (Platurus) has the ventral shields so much developed as to be able to move on land. No Oceanic Serpent is known of gigantic dimensions, such as is currently alleged to have been seen by unscientific observers.
"Although these five groups," remarks Dr. Günther, "are not separated from each other by defined lines of demarcation, and frequently pass into one another by intermediate forms, yet a family and genus which should be composed of species of several of these groups would be a very unnatural assemblage of heterogeneous forms."
It is also remarked by the same naturalist that there is no sharp boundary line between the order of Snakes and that of Lizards. There are various limbless Saurians of Ophidian appearance, but the systematic position of which is decided by the structure of their jaws. The Common Orvet, or Slow-worm, is a familiar instance. On the other hand, certain Ophidians remind us, by several characters, of the Saurian type,—as the Snakes constituting the families Typhlopidæ, Tortricidæ, Xenopeltidæ, and Uropeltidæ, which are distinguished by polished, closely adherent, rounded, sub-equal scales, much resembling the smooth scales of various Scincoid Lizards; most of them have a very narrow mouth, unlike the enormous gape of the typical Serpents, and some are without that longitudinal fold in the median line of the chin which is so characteristic of most Ophidians; moreover, most of them have rudiments of the bones of a pelvic arch. "The reason," alleges Dr. Günther, "why we adopt the view of those systematists who refer such reptiles to the Ophidians, instead of associating them with the limbless Scincoid Lizards, is the loose connection of the jaw-bones, a character which must be considered as peculiar to the Ophidians, and which is only somewhat less developed in the families mentioned than in the typical forms. The two halves of the lower jaw in Ophidians, namely, are not united by a bony symphysis, but by an elastic ligament. The peculiar mobility of the jaw bones enables the Snakes to extend the gape in an extraordinary degree, and to work their prey down through the collapsed pharynx."
The same naturalist classifies the Ophidia into three sub-orders, in which the venemous Snakes are separated from the others; but to some herpetologists this arrangement must appear rather forced, as his Venemous Colubrine Snakes have certainly a much nearer resemblance in other respects to the Colubridæ than they have to the Viperine Snakes. For the most part, these reptiles are provided with numerous teeth, which are lengthened, conical, thin and pointed like a needle, and more or less bent backwards.
In Dr. Günther's first sub-order, that of Non-venemous Snakes, the teeth are either entirely smooth, or only the last of the maxillary series is provided with a faint longitudinal groove, which is not intended to convey a virus into the wound, the groove appearing rather to increase the strength of the tooth. Many of them have long teeth in front of the jaws or of the palate, but these are never grooved or perforated, and only serve to afford a firmer hold on the living and struggling prey.
"The structure of the venom-tooth is not the same in all poisonous Snakes: in some it is fixed to the maxillary bone, which is as long or nearly as long as in the non-venemous Snakes, and generally bears one or more ordinary teeth on its hinder portion. The venom-tooth is fixed more or less erect, not very long, and its channel is generally visible as an external groove. The poisonous Snakes with such a dentition have externally a more or less striking resemblance to the non-venemous Serpents, and on this account they are designated as Venemous Colubrine Snakes, forming our second sub-order." Two very distinct families are here brought together—viz. the Elapidæ (which comprises the Cobras and many others), and the Hydrophidæ (or Sea Snakes).
"In the other venemous Snakes, composing the third sub-order, the maxillary bone is extremely short, and does not bear any teeth except an exceedingly long fang, with a perfectly closed externally invisible channel in its interior. Although this tooth also is fixed to the bone, the bone itself is very mobile, so that the tooth, which is laid backwards when at rest, can be erected the moment the animal prepares to strike. This tooth or fang, like all the other teeth, is not only occasionally lost, but appears to be shed at regular intervals. From two to four other venom-fangs in different stages of development, destined to replace the one in action, exist between the folds of the gum, and are not anchylosed to the bone." The more characteristic venemous Snakes appertain to this sub-order—viz. the two families Crotalidæ (comprehending the Rattlesnakes, the Fer-de-lance, &c.) and Viperidæ (comprising the Vipers, Puff-adders, &c.).
Let it be particularly borne in mind that the supposed distinguishing characters of all poisonous Snakes, as assigned by sundry mischievously ignorant writers, are those of the third of the foregoing sub-orders almost exclusively. Even the broad, flat, and lanceolate form of head is exemplified in certain Tree Snakes of the non-venemous genus Dipsas, and not in the Cobras and others that are quite as deadly—e.g. Hoplocephalus, Bungarus, Naja, Elaps, and others constituting the Colubriform family Elapidæ.