Home

Once upon a time there was a thick textbook on vertebrate anatomy, and its second volume was...

Dietrich Starck: Vergleichende Anatomie der Wirbeltiere auf evolutionsbiologischer Grundlage, 2: Das Skeletsystem [sic]: Allgemeines, Skeletsubstanzen [sic], Skelet [sic] der Wirbeltiere einschließlich Lokomotionstypen, Springer 1979

... and on p. 619 to 622, it had the following to say on bird hand embryology (my translation). Keep in mind that the author took for granted that birds were not dinosaurs but, you know, "descended from the thecodontians", and also took trees-down for granted:

        For a long time it was in dispute to which fingers the three rays of the bird hand correspond. Embryological and paleontological finds have shown that the remaining fingers correspond to the rays I, II and III (fig. 472 [below]). In the embryonic bird hand all five rays are present (Steiner). The Anlage of the bird hand is identical to the embryonic reptile hand. The radius Anlage lays at an angle to the ulna Anlage and continues into a precarpal (prepollex). The fourth finger is the most strongly developed one in the early embryonic hand. Metacarpal IV is preserved for a long time and becomes cartilaginous. It can probably be incorporated into the ossification of metacarpal III. The fifth digital ray is clearly visible in embryos of Anas and Anser on the 6^th – 7^th day of brooding (fig. 472). It stays blastematous, however, and does not chondrify at first. The basal [ = proximal] part of the 5^th ray undergoes a secondary development and is incorporated into the carpal skeleton.

        The definitive carpal skeletal elements of the birds do not correspond to the primary carpalia (Steiner). The primarily initiated [angelegt – related to Anlage] show the typical tetrapod pattern. The intermedium branches off of the ulnar ray and continues into the I^st and II^nd finger. The radiale, two centralia and the ulnare have been shown to exist. The intermedium completely regresses early on. The radiale and the centrale radiale fuse and form the radial carpal element of the definitive wing (fig. 473 [below]). The centrale ulnare and the ulnare regress as well. Four distal carpals occur in ontogenesis. First distal carpal 1 fuses to 2 and 3 to 4. Carpals 1 and 2 then fuse to the proximal end of metacarpal I and disappear in it. Carpals 3 and 4 join metacarpal III. In the end the united skeleton piece fuses to metacarpal II to form the so-called os metacarpi. The basal part of the fifth ray is preserved and becomes larger. It moves to the palmar side from the ulnare, chondrifies and then ossifies and becomes the ulnar carpal element of the definitive carpus (fig. 473). These changes of the primary carpus of the birds can be directly read from ontogenesis. The old interpretation of the radial carpal bone as an intermedio-radiale is thus useless. The disappearance of the intermedium is initiated early in the archosaur stem (†Protorosaurus). The fusion of the distal carpal elements to at first two bones is likewise present in ancestral archosaurs. The reduction of the 4^th and 5^th fingers, too, begins long before the gain of flight ability. In the Dinosauria similar changes as in the birds have evolved independently from the bird stem in parallel evolution. There, too, the first three fingers are much more strongly developed than the 4^th and 5^th. The fifth ray is reduced, but its basal part is short and thick in †Thecodontosaurus and †Plateosauridae [sic] and begins a new differentiation. The fusion of radiale and centrale radiale likewise starts soon among reptiles. The phylogenetic development thus proceeds in the same way as the ontogeny of the skeleton in the bird wing. In †Archaeopteryx (fig. 473 b) the reduction of the 4^th and 5^th finger is already reached. Finger I is strongest, finger II longest. [This rhyme doesn't occur in the original.] Metacarpalia I to III are still free. [This one does.] The number of phalanges still is 2-3-4 in †Archaeopteryx, like in primitive forms. It is questionable if a rudiment of the 4^th metacarpal is preserved. The carpus is not completely preserved in the remains of †Archaeopteryx. A distal piece is interpreted as a fusion product of the distal carpals. Remains of the fused radiale and centrale radiale are likewise preserved. Whether a free ulnare still existed is unknown.

Fig. 472. [Click on it to see it in full resolution and very large size.] Blastematous, [and?] cartilaginous Anlage of the hand skeleton of goose embryos. a 7½ days; b 8½ days old. (Redrawing after Steiner, 1934)

1 prepollex; 2 centrale ulnare; 3 centrale radiale; 4 radiale; 5 radius; 6 ulna; 7 intermedium; 8 ulnare; 9 metacarpal V; 10 [distal] carpal 4; 11 metacarpal IV; 12 [distal] carpal 3; 13 [distal] carpal 1; I – V digital rays

Fig. 473. [Click on it to see it in full resolution and very large size.] a Hand skeleton of a bird embryo (Sterna, tern) at incipient ossification. b Right arm skeleton of the primordial bird, †Archaeopteryx. (a after Wiedersheim; b after Jaekel out of Versluys)

1 humerus; 2 radius; 3 ulna; 4 ulnare [true ulnare in b, "metacarpal V" in a]; 5 intermedium; 6 radiale; 7 [distal] carpal 1 + 2; 8 metacarpal I; 9 metacarpal III; 10 metacarpal IV; 11 metacarpal V; 12 [distal] carpal III; 13 metacarpal II; I – V digital rays

Now for the interpretation.

The figures 472 a and b don't show an actual prepollex, but they leave the space for it distal to the radiale. In exactly this space Feduccia & friends have repeatedly found an apoptotic digit that they call the first. I don't see why it shouldn't be a prepollex. The blastematous spur interpreted as the 5th finger is in a place where it may also be visible in one photo by Feduccia & Nowicki and where Nikbakht & McLachlan (1999, fig. 3 (e)) were able to let some pointed cartilage grow when they implanted an FGF-4-soaked bead into the caudal edge of an embryonic chicken wing.

In the same figures the 5^th finger is proximal, not caudal, to the 4^th; this is the normal state of affairs in dinosaurs (other than, basically, sauropods) where that digit hasn't disappeared.

According to the text the 4th finger is the most easily visible one in early stages of development. This fits the interpretation by Feduccia and AFAIK everyone else, and it also fits the interpretation that the 4^th is the longest in adult amniotes generally (unless modified by evolution) ( = plesiomorphically). The implication by the description of the later stages, however, implies that IV simply stops growing when I to III inflate to occupy the entire breadth of the hand (and fig. 472 looks a lot like just this). This would mean that the digit called IV by Feduccia & friends in the early stages is not the same as the digit called IV by them in the later ones. It also reminds me of the small nubbin of bone on the caudopalmar side of the proximal end of metacarpal III in Sinraptor (a basal carnosaur, thus a close relative of Coelurosauria) that is interpreted as metacarpal IV. It also seems to be what Nikbakht & McLachlan (1999) call "element 5".

That the intermedium continues into the 1^st and 2^nd rays is not something I'll accept at face value; digital rays actually growing out of the intermedium (instead of the ulnare or, in the case of the prepollex, the radiale) are AFAIK a peculiarity entirely unique to salamanders. Probably all that is meant is that these rays lie distal to the intermedium (which is the case, especially for II).

To mention the basal archosauromorph Protorosaurus for the complete disappearance of the intermedium "early in the archosaur stem" is misleading. Allosaurus, for example, still has one (though small) (Chure 2001). Tyrannosaurids and ornithomimids are also reported to retain it.

The fusion of the radiale with the centrale immediately distal to it is not surprising either. Herrerasaurus has a small, round centrale (its only centrale) distal to the radiale, while theropods lack discernible centralia entirely. (Feduccia, 1996, was right in saying that theropods have lost an entire row of carpals – the middle one!) I should find out whether the same holds for Sauropodomorpha (in which case the fusion of the last centrale to the radiale could be an autapomorphy of Eusaurischia, the clade that includes Theropoda and Sauropodomorpha but excludes Eoraptor and Herrerasauria).

The disappearance of the ulnare is thoroughly weird, but AFAIK it is universally accepted. In keeping with this interpretation, Allosaurus has no trace of an ulnare even in well-preserved, articulated hands. (It has four carpals, interpreted as a big radiale, a big distal carpal 1, a small intermedium and a very small distal carpal 2 by Chure (2001). Most of metacarpal II seems to articulate directly with the ulna! Metacarpal III may not reach the carpus at all, if this isn't due to some disarticulation as shown for Deinonychus by Gishlick (2001).) Neither does Deinonychus (Gishlick, 2001; also see below). Reports of ulnaria in a few other coelurosaurs, such as Caudipteryx, could be misinterpretations of intermedia, though this remains to be tested. (Or at least I need to have a good look at some primary literature... :-) )

The fusion of distal carpals 1 and 2 is common in theropods; all that have ever been suggested as the closest relatives of the birds have it. (Some of the rest may be explicable by reversals. Segnosaurs have a suture between these two bones.)

The ulnare is said to be replaced by "the basal part of the fifth ray" which "moves to the palmar side". In Herrerasaurus and other non-theropodan saurischians metacarpal V is famously palmar to IV (which is palmar to III, which in turn becomes palmar to II in Maniraptora or maybe earlier). The observation that it moves to the palmar side and replaces the ulnare identify it as the famous "element X", which was recently found to express a gene that is also expressed in digital rays but not in proximal carpals (Welten et al. 2005). – This seems to be what Nikbakht & McLachlan (1999) call a "pisiform". Does a pisiform occur anywhere outside of mammals??? (And what, actually, is it? A sesamoid? The infamous "postminimus"?)

In Archaeopteryx of course no trace of a 4^th metacarpal is preserved, and there really seem to be only two carpals in total, which are most probably the radiale and dc 1 + 2 as in Deinonychus (Gishlick 2001) in which likewise only two carpals of similar shapes to those of Archaeopteryx are preserved. (Ostrom originally interpreted the distal carpal of Deinonychus as the radiale, even though it articulates with all three metacarpals. Inexplicably this rather odd interpretation was taken at face value and reproduced by Feduccia (1996).)

In sum, even though the picture isn't absolutely ideally neat (it is interesting instead :-) ), I don't see any reason why the fingers of adult birds should not be regarded as I, II and III, and I don't see any reason not to regard birds as theropods.

Sure, line drawings redrawn from a work from 1934 don't instill a lot of confidence in me. But apart from the finding of the two rather ephemeric centralia (which might be hidden in some of the photos by Nikbakht & McLachlan) and the absence of the relatively elusive prepollex, the drawings depict the exact same things as color photos derived from state-of-the-art staining methods in the last 10 years. Even things like the movement of "element X" are there. Only the labels are different.

Daniel J. Chure: The wrist of Allosaurus (Saurischia: Theropoda), with observations on the carpus in theropods, p. 283 – 300 in Jacques Gauthier & Lawrence F. Gall (eds): New Perspectives on the Origin and Early Evolution of birds: Proceedings of the International Symposium in Honor of John H. Ostrom, Peabody Museum of Natural History/Yale University 2001

Alan D. Gishlick: The function of the manus and forelimb of Deinonychus antirrhopus and its importance for the origin of avian flight, p. 301 – 318 in the same book

Neda Nikbakht & John C. McLachlan: Restoring avian wing digits, Proceedings of the Royal Society of London B: Biological Sciences 266, 1001 – 1004 (7 July 1999)

Monique C. M. Welten, Fons J. Verbeek, Annemarie H. Meijer & Michael K. Richardson: Gene expression and digit homology in the chicken embryo wing, Evolution & Development 7(1), 18 – 28 (January 2005)

Spambot, terminate yourself.