Among living tetrapods, birds are unique in having completely separated the locomotor functions of fore and hindlimbs. The propulsive excursions of the forelimbs, which primarily involve elevation and depression in a transverse plane, differ fundamentally from those of most other tetrapods (pterosaurs and bats excepted) in which the forelimbs protract and retract in anteroposterior planes.
Pterosaurs and birds present a number of striking parallelisms in the structure of their flight apparatus and the glenoid is yet another example of their independent derivation of similar features.
In both rhamphorhynchoid and pterodactyloid pterosaurs the glenoid is distinctly saddle shaped with laterally as well as dorsally facing regions of the articular surface.
The origin of the pterosaurian glenoid must have involved the same evolutionary migration of position and orientation that has been outlined here for the avian lineage.
In contrast to the bulbous humeral head of birds, however, the humerus of pterosaurs bears a saddle-shaped facet, thus constraining the wingbeat excursion. This difference is likely a reflection of the relative structural versatility of the two wing types: an outstretched, sail-like membrane supported principally by a single digit versus a flexible airfoil composed of individual feathers, each with its own structural and functional integrity.http://books.google.ca/books?id=8CKYxcylOycC&pg=PA243&lpg=PA243&dq=glenoid%20fossa&source=bl&ots=SopV9CAGec&sig=-gWOltWiFGplrU9tcXV8X8pBPUI&hl=en&ei=yTjS5aEBoT6lwf_uMC9Ag&sa=X&oi=book_result&ct=result&resnum=6&ved=0CC8Q6AEwBQ#v=onepage&q=glenoid%20fossa&f=false
From the article on page 267 (by Frey et al.):
As in birds, the glenoid fossa in most pterosaurs is elevated by a dorsolaterally directed elongation of the coracoid and lies almost level with the vertebral column
The [pterodactyl pterosaur] coracoid is about 75 per cent of the length of the scapula. It is expanded at its contact with the scapula, but has a more gentle decrease in width over its length. A small, blunt coracoid process is present, but it is not possible to tell if a groove separates it from the glenoid fossa. The sternal articulation is concave, faces posteroventrally, and lacks a posterior expansion. A large glenoid fossa faces anterodorsally with a dorsoventrally concave and anteroposteriorly convex saddle shape.
Wing skeleton. Both [pterodactyl pterosaur] wings are present in NGMC 99-07-1 (Text-figs 2, 4; Table 2). The humeri are complete though the right deltopectoral crest has become detached and rotated from its anatomical position (Text-fig. 2). The humeral head has an anteroposteriorly concave and dorsoventrally convex, saddle-shaped articulation so that it mirrors the shape of the glenoid.
Whether or not Microraptor could achieve powered flight or only passive gliding has been controversial. While most researchers have agreed that Microraptor had most of the anatomical characteristics expected in a flying animal, some studies have suggested that the shoulder joint was too primitive to have allowed flapping. The ancestral anatomy of theropod dinosaurs has the shoulder socket facing downward and slightly backward, making it impossible for the animals to raise their arms vertically, a prerequisite for the flapping flight stroke in birds. Some studies of maniraptoran anatomy have suggested that the shoulder socket did not shift into the bird-like position of a high, upward orientation close to the vertebral column until relatively advanced avialans like the enantiornithes appeared. However, other scientists have argued that the shoulder girdle in some paravian theropods, including Microraptor, is curved in such a way that the shoulder joint could only have been positioned high on the back, allowing for a nearly vertical upstroke of the wing. This possibly advanced shoulder anatomy, combined with the presence of a propatagium linking the wrist to the shoulder (which fills the space in front of the flexed wing and may support the wing against drag in modern birds) and an alula or "bastard wing" may indicate that Microraptor was capable of true, powered flight.
It is not an easy task to get all the needed information about the shoulder joint but this is how it appears:
Rhamphoryncidae had a saddle joint. Both the glenoid fossa and the humerus head were saddle-shaped.
Basal paraves - glenoid fossa was still saddle shaped but the humerus head was bulbous.
PTEROSAUR scapula, coracoid and glenoid
Some pterosaur bones are quite unusual. This scapulo-coracoid is photographed from both sides. The glenoid cavity of the shoulder joint can be seen, where the humerus articulates the wing to the body.
A large [pterodactyl] glenoid fossa faces anterodorsally with a dorsoventrally concave and anteroposteriorly convex saddle shape.
Scapula orientation in theropod dinosaurs is quite interesting and it is worth looking, to begin with, at what orientation is displayed in primitive reptiles. The scapula is generally held at an angle of 90 degrees to the horizontal line held by the backbone – in other words it was held in a perpendicular fashion. At the other extreme, extant birds rotated the scapula so that it lies parallel to backbone – a position also evolved by the pterosaurs.
Theropods, and non-avian dinosaurs in general (but not bird-like theropods), evolved a condition that can be described as something in between – an intermediate position if you will. The scapula is held in an oblique position laterally to the ribcage but actually determining the exact position is somewhat problematic. There are not that many fully articulated specimens that can be referred to and there is always the spectre of both taxanomic and taphonomic variation to throw yet another spanner into the works.
Posteriorly facing glenoid fossa
Agnolín and Novas. 2013. Avian ancestors
In this way, the scapulae of unenlagiids lie close to the vertebral column, dorsal to the ribcage, with the flat costal surface of the scapular blade facing ventrally, a condition seen in microraptorans (i.e. Microraptor), basal avialans (e.g. Archaeopteryx, Rahonavis), and ornithothoracine birds (Senter 2006), in which the shoulder socket sits high on the back, and the margins of the glenoid are smooth, thus this surface becomes shalower and consequently more continuous with the rest of the lateral surface of scapula
(Burnham 2008). In sum, the lateral orientation of the scapular glenoid in unenlagiids
(and probably also in other basal averaptorans), together with the absence
of acute ridges delimitating the glenoid cavity, suggest that the humerus in these
taxa was able to be elevated close to the vertical plane, as proposed by Novas and Puerta (1997) (Figs. 5.1, 5.2).
It is important to mention that scansoriopterygids retained a caudoventrally oriented glenoid, a subrectangular coracoid with reduced biceps tubercle, and a distally fan-shaped scapular blade, all representing plesiomorphic character states in respect to paravians.=======================================================
Here is a good overview of the shoulder girdle of modern birds:
1. Sternum / breastbone 2.Coracoid 3.Clavicles / furcula 4. Scapula 5.Joint with the wing 6.Foramen trioceum
Here is a very interesting video: