Examples from Elgin, Hone and Frey (in press) which they purport to demonstrate the wing trailing edge attaching to the ankle. In fact, none of these examples do so. Note: Hone's blog: http://pterosaur-net.blogspot.com/2010/09/on-wing-membrane-and-ankle-attachment.html exchanged many of the images since the original post.
A. the darkwing specimen of Rhamphrohynchus (JME SOS 47484). Note the trailing edge is directed toward the elbow, as in the Peters (2001) model and the Zittel wing. It does not curve toward the ankle. The authors missed the curve inboard of the elbow. The pink area is on a lower plane.
B. Jeholopterus (IVPP V 12705). The authors failed to use the counterplate and traced the proximal membrane extending to the elbow and the wingtip attaching to the ankle!!
C. The “Vienna specimen” (NHMW 1975/1756) of Pterodactylus. Here the proximal attachment points could not be more clear, so the authors invented "membrane shrinkage" to explain the reduction of the membrane near the elbow. Notice the rest of the wing membrane failed to shrink.
D. A new specimen attributed (Bennett 2007) to Anurognathus. Here again the authors traced the trailing edge directly to the elbow without a tibia connection.
E. Eosipterus (GMV 2117). The only portions of the patchy wing membrane traced by the authors followed the Peters (2001) model precisely with no indication of a membrane lateral to the tibia
F. Sordes (PIN 2585/3). The authors did not understand that the medial trailing edge was torn from the antebrachium, which forms the trailing edge leading toward the ankle, the continues on as an apparent fiber-embedded uropatagium. Details here. No other pterosaur uropatagium has aktinofibrils. The actual uropatagia are smaller, matching those of other pterosaurs.
G. An unnamed specimen (SMNK PAL 3830) in which a swath of torn and disassociated wing membrane crosses both legs at an angle. This specimen provides no data as to how this membrane was connected to anything on the pterosaur, including either tibia. However the membrane itself does not extend beyond a narrow rectangular band across its entire length, matching the Peters (2001) hypothesis, not the deep triangle of tissue imagined by Elgin, Hone and Frey (2010).