The evolution from primitive, asymmetrical, reptile hand and foot to a more symmetrical therapsid hand and foot has been widely recognized (Fig. 1). Primitively the phalanges were more or less subequal and, as a consequence, the digits increased in length from 1 to 4. The primitive pattern is found on sprawling tetrapods, like Gephyrostegus (Fig. 2). It permits the digits to avoid hitting the substrate during the recovery stroke of the sprawling limbs, while maximizing digit length.
In basal therapsids, like Biarmosuchus (Fig. 1), the reduction of phalanges 3.2, 4.2 and 4.3 and the resulting increased symmetry in the manus and pes tells us the limbs were not so sprawling any more. The recovery stroke occurred more below the knees and elbows, rather than out to the sides. Later, about the time certain therapsids evolved into mammals, these three disc-like phalanges ultimately disappeared.
Pterosaurs undertook a similar evolution in the pes (Fig. 3), in which phalanges 3.2, 4.2 and 4.3 were the shortest phalanges in those digits in derived forms. The pattern of reduction is both distinct and convergent in several pterosaur lineages. Each pattern produces a distinct foot shape that should aid in the identification of trackmakers (Peters 2011) when examining potential pterosaurian ichnites. In no pterosaur did the short phalanges ultimately disappear.
The interesting thing is the evolution of the pterosaurian pes did not match the evolution of the gait. Rather, just the opposite. Early pterosaurs (Dimorphodon) had a right angle femoral head and an upright stance, but they retained a more lizardy foot with digits of increasing length laterally. Derived pterosaurs (Pteranodon) had a more sprawling femur (judging by the angle of the head to the shaft) and a more symmetrical pes with short mid phalanges on digits 3 and 4. Not sure of a good explanation for this yet.
Then there’s the subject of pterosaur clawed fingers. In MPUM 6009 the (non-wing) finger phalanges were more or less subequal. In the ornithocheirid pterosaurs, Brasileodactylus and Arthurdactylus, m3.2 was a mere disc. In these pterosaurs, manual digits 2 and 3 were subequal in length as a consequence. Not sure why yet, but these pterosaurs had such long proximal wing elements set so far out in front of their feet that they were not able to transfer weight to them while keeping their feet beneath the humeral glenoid, the center of balance in all pterosaurs. It also seems unlikely that in ornithocheirids the wing fingers could have had much contact with the substate sitting on top of the largely horizontal manus and wing finger. Such a pattern is unlike that of typical trackmakers like Pterodactylus and Ctenochasma (Fig. 4), which had shorter and more gracile proximal wing elements.
In Istiodactylus the manus phalangeal formula for digits 1 through 3 was 2-2-2. So there was some fusion and loss here, the only case of seen of it in pterosaurs.
As always, I encourage readers to see specimens, make observations and come to your own conclusions. Test. Test. And test again.
Evidence and support in the form of nexus, pdf and jpeg files will be sent to all who request additional data.
Peters D 2011. A Catalog of Pterosaur Pedes for Trackmaker Identification. Ichnos 18(2):114-141. http://dx.doi.org/10.1080/10420940.2011.573605