In dividing and grouping such teeth for the purpose of description and comparison, more definite characters are required than those derived merely from form or function. The first step towards a classification has been made by the observation that the upper jaw is composed of two bones, the premaxilla and the maxilla, and that the suture between these bones separates the three anterior teeth from the others. These three teeth, then, which are implanted by their roots in the premaxilla, form a distinct group, to which the name of “incisor” is applied. This distinction is, however, not so important as it appears at first sight, for, as mentioned when speaking of the development of the teeth, their connection with the bone is only of a secondary nature, and, although it happens conveniently for our purpose that in the great majority of cases the segmentation of the bone coincides with the interspace between the third and fourth tooth of the series, still, when it does not happen to do so, as in the case of the Mole, we must not give too much weight to this fact, if it contravenes other reasons for determining the homologies of the teeth. The eight remaining teeth of the upper jaw offer a natural division, inasmuch as the posterior three never have milk-predecessors; and, although some of the anterior teeth may be in the same case, the particular one preceding these three always has such a predecessor. These three then are grouped apart as the “molars,” or, since some of the teeth in front of them often have a molariform character, “true molars.” Of the five teeth between the incisors and molars the most anterior, or that which is usually situated close behind the premaxillary suture, almost always, as soon as any departure takes place from the simplest and most homogeneous type, assumes a lengthened and pointed form, and is the tooth so developed as to constitute the “canine” or “laniary” tooth of the Carnivora, the tusk of the Boar, etc. It is customary therefore to call this tooth, whatever its size or form, the “canine.” The remaining four are the “premolars” or “false molars.” This system of nomenclature has been objected to as being artificial, and in many cases not descriptive, the distinction between premolars and canine especially being sometimes not obvious; but the terms are now in such general use, and are so practically convenient—especially if, as it is best to do in all such cases, we forget their original signification and treat them as arbitrary signs—that it is not likely they will be superseded by any that have been proposed as substitutes for them.