The five groups of snakes described in chap. ii. are divided into three sub-orders of Ophidia as follows:—1. _Ophidia colubriformes_ (the harmless snakes). 2. _Ophidia colubriformes venenosi_ (those which, not having the viperine aspect just now described, are the more dangerous from their innocent appearance). 3. _Ophidia viperiformes_ (the viperine snakes). Although apparently named from their form only, it is the teeth which have chiefly to do with these latter distinctions, as will be seen on reference to the dotted examples of upper jaws. The first have the six rows of simple teeth (four above, as seen, and the lower jaw teeth), in all from 80 to 100 perhaps. The second have the two rows of palate teeth, the lower jaw teeth, and a _fixed_ fang on each upper jaw, with one, two, or more simple teeth in addition. The Australian poisonous serpents are nearly all of this group, the only viperish-looking one, the ‘Death adder,’ having fixed fangs like the cobras. The sea snakes and the _Elapidæ_ are included. The third have only four rows of simple teeth, viz. those of the lower jaws and those of the palate, with a solitary moveable fang in each upper jaw. [Illustration: From Fayrer’s _Thanatophidia_. The four larger dots represent fangs.] Fayrer divides the poisonous snakes of India, again, into four families, viz. _Elapidæ_ and _Hydrophidæ_, with fixed fangs; and _Viperidæ_ and _Crotalidæ_, with mobile fangs. But without so many perplexing distinctions, I hope to be able to interest the reader in that wonderful piece of mechanism, the poison fang, and by the aid of the authorities to represent it in simple language. We have long been accustomed to read that a serpent’s fang is a ‘perforated tooth’ or a ‘hollow tube,’ as if a miniature tusk had a hole bored through its entire length, the poison entering at the root and flowing out again at the point. This is not strictly the case. Fangs in their construction are not absolutely ‘hollow,’ with ivory on the outside and pulp on the inside, but are as if you had flattened out an ivory tusk and folded or wrapped it over again, so as to form a pointed tube. It would then have dentine both on the outer and inner surface. This involution may be compared with that seen in a long narrow leaf, in which the larva of an insect has enwrapped itself. The various degrees of involution are extremely close, as also would be the forms of leaves and the extent of curling which each caterpillar had effected. Some fangs are folded so as to leave the—_join_, we will call it, easily perceptible. Others leave a groove more or less evident; while in others the fold is so complete as to have disappeared entirely. Schlegel, in describing the insensible passage from solid teeth to fangs, affirms that traces of the groove are always perceptible: ‘_On découvre toujours les traces de la fente qui réunit les deux orifices pour le venin._’[94]