By digitizing the enamel configuration and calculating the overall directional tendencies of the enamel edges in complex equid dental patterns, quantitative changes in functional attributes were traced through a phylogenetic series. Pronounced changes both in the directional emphasis of the enamel edges and in the chewing motion occurred and were interrelated. Two distinct trends were found. From early Eocene to Oligocene, edge alignment became increasingly confined to a single direction, but the teeth retained the phase I and phase II chewing angles characteristic of many primitive herbivorous and omnivorous mammals. The chewing direction in the horizontal plane also remained essentially unchanged. During the Oligocene and the early Miocene, both edge direction and chewing direction remained stable. However, the middle Miocene genus Merychippus represents a stage in which the enamel directional emphasis was lost in transition to a new anteroposterior emphasis, a trend that continued into the late Tertiary equine taxa. Concurrent with this transformation was a loss of all but vestiges of the phase I and II chewing angles in the vertical plane and a shift of the chewing direction in the horizontal plane to an almost transverse movement of the jaw.
In the Oligocene equids, the edges of the labial cusps did attain perpendicularity to the phase I chewing direction, which was oblique to the horizontal plane, but the chewing direction in the horizontal plane was far from perpendicular to the lingual enamel edge direction because the lingual cusps were configured for efficiency in compressive, not shearing, motion that dominated the phase II surfaces. In the later Tertiary the chewing direction and edges attained essentially a perpendicular relationship, which represents the theoretical optimum for a shearing system. The radical change in edge direction in the Miocene probably occurred as a result of an increase in translative movement (longer glide) between the surfaces, making the lingual edges subject to selection for perpendicularity and flattening of the occlusal surface, resulting in a condition where a single edge direction became optimal for all cusps. The extended chewing stroke translation was perhaps the key event initiating the other changes and would have resulted from increased chewing effort accompanying a shift to a diet of mainly grass. The loss of the dual (shearing and compressive) function of the occlusal surface suggests that the earlier diet may have contained significant proportions of other plant parts, such as fruits, seeds, or tender leaves that might be optimally chewed by compressive movements.