The Miocene ape Nacholapithecus is known from rather complete skeletons; some of them preserve the shoulder joint, identified by three scapulae and one clavicle. Comparisons made with other Miocene and living apes ( Proconsul, Equatorius, Ugandapithecus) suggest that the mobility of the scapulohumeral joint was important, and scapular features such as the morphology and position of the spine and the morphology of the acromion and axillary border resemble those of climbing arboreal primates except for chimpanzees, gorillas, or orang-utans. From the size of the scapula (male Nasalis size), it is clear that the animal is smaller than an adult chimpanzee, but the clavicle is almost as relatively long as those of chimpanzees. Some features closer to colobine morphology reinforce the hypothesis that Nacholapithecus was probably a good climber and was definitely adapted for an arboreal life.

Intergeneric morphological variation of the glenohumeral joint surface was investigated among three ateline genera (Ateles, Lagothrix, Alouatta) and compared with Cebus (an ancestral morphotype of atelines) and Hylobates (a specialized brachiator) to reveal characters associated with forelimb suspensory behavior. Seventy-six skeletal specimens were examined, and articular surface curvature was measured by a three-dimensional digitizer. It was found that Ateles exhibits joint features distinct from the other atelines, but resembles Hylobates in its large breadth-length ratio of the glenoid surface and the humeral head, a relatively spherical humeral head, and a dorsoventrally extensive humeral head relative to the glenoid surface. These morphologies are likely to be related to brachiation, rather than to climbing behavior. A dorsoventrally extensive glenohumeral joint is interpreted to facilitate an increased stride length during brachiation. Lagothrix was found to show many primitive features that are shared with Alouatta in spite of its forelimb suspensory behavior. This may be related to the less specialized mode of forelimb suspensory behavior in Lagothrix compared with Ateles. Those characters that apparently correspond to dependency on suspensory behavior can be useful in interpreting the positional behavior of extinct primate taxa.

An elongated clavicle is one of the distinct features of apes and humans. It plays an important role in providing mobility as well as stability for the shoulder joints. The relative length of the clavicle is an especially important factor in limiting the range of shoulder joint excursion. It is said that among primates, Asian apes, i.e., gibbons and orang-utans, have very long clavicles. At the same time, they also have a wide upper thoracic cage, which may diminish the effective length of the clavicle. To clarify the length of the clavicle in apes, from the standpoint of the functional anatomy of the shoulder girdle, we examined clavicular length in 15 anthropoid species exhibiting various positional behaviors. The results confirm that clavicle length in Asian apes is long, and chimpanzees have a short clavicle like that of Old and New World monkeys, when scaled to body mass. The clavicular length of chimpanzees, however, is intermediate between Old World monkeys and Asian apes when scaled against thoracic width. Therefore, living apes can be grouped together, albeit just barely, by possession of a relatively long clavicle for their thoracic cage size. Interestingly, New World monkeys tend to exhibit a longer clavicle than Old World monkeys of equivalent body mass or thoracic cage width. Although it is unclear whether the ancestral condition of clavicular length in anthropoids was similar to that of living Old or New World monkeys, an elongation of clavicle was an important step toward evolution of the modern body plan of hominoids.

While a relatively broad thorax and strongly curved ribs are widely regarded as common features of living hominoids, few studies have quantitatively examined these traits by methods other than calculating the chest index. The present study aims to quantify variations in thoracic cage morphology for living anthropoids. The odd-numbered ribs (first to eleventh) were articulated with the corresponding vertebrae and the cranial and lateral views subsequently photographed. Rib profiles were digitized in both views and line-fitted by a Bézier curve to create a three-dimensional morphological data set. When thoracic cage width was scaled against body mass, Hylobates (and possibly Pongo) plotted above non-hominoid anthropoids at almost all rib levels, while Pan did not differ from non-hominoid anthropoids. The overall pattern of the normalized thoracic width differed between Hylobates and other hominoids. In Hylobates, an upward convex curve was seen between the first and seventh ribs while a more linear pattern was observed in Pan and Pongo. This result quantitatively confirmed that the barrel-shaped thoracic cage in Hylobates can be distinguished from the funnel-shaped form in other hominoids. Conversely, all hominoids shared two distinct features in the upper half-thorax: (1) a pronounced dorsal protrusion of the proximal part of the rib in accordance with ventral displacement of the thoracic spine and (2) a relatively medially projecting sternal end. Although these features are likely to provide some mechanical advantage in orthograde and/or suspensory positional behaviors, they were barely present in the suspensory Ateles.