Illustrated systematic descriptions of normal primate dentitions are given by James (1960) and Swindler (1976; 1978).

For obvious reasons, much less is known about variations in the roots of teeth than in the crowns. However, it is well-established that in man the roots of the teeth, particularly the molars, are no less variable than the crowns. This variability is least in M1 but is markedly greater in M2 and greater still in M3. The M2 quite frequently have extra rootlets and some degree of fusion between roots is also common. M3 are much more affected in this way and, in addition, the overall length of their roots is very variable. Descriptions of the roots of the teeth of the non-human primates, for example Bennejeant (1936), suggest that root morphology is probably as variable as in man and that the types are similar, variability being greatest in M3. Bennejeant (1936, pp.115, 182) drew attention to longitudinal grooves on the labial surfaces of the roots of some primate upper incisor teeth and less commonly on upper canines. An example in a baboon is shown in Figure 20.5. These grooves cannot be regarded as a tendency for the formation of two roots because the upper incisors and canines never have two roots. The grooves could indicate minor degrees of connation or dichotomy.

In man, M1 often has a small cusp, the cusp or tubercle of Carabelli, on the lingual aspect of the protocone in addition to the normal four. If the cusp is well-developed, it is usually present on m4, M2 and M3 as well. The prevalence of the cusp on M varies between different racial groups (Sofaer, 1981, p. 153).