Heilmann, Thompson, Beebe, Tetrapteryx and the Proavian

Many people familiar with ideas on the early evolution of birds and of bird flight will know of the Proavis or proavian, a hypothetical bird ancestor illustrated and discussed by William Pycraft (1868-1942) in 1906 but made better known by Gerhard Heilmann (1859/1861-1946) during the 1910s and 20s. Versions of this creature were later illustrated in colour by Zdeněk Burian for popular books of the 1960s, 70s and 80s. In the article that you’re about to read, we’re predominantly interested in Heilmann’s take on the Proavis.

Fighting, gliding and climbing proavians, depicted in a cluttered woodland environment, by Gerhard Heilmann. This is typical of Heilmann’s many excellent scenes. He used posed plaster models of the creatures before drawing. This illustration was used in the 1916 Danish version of the book (it first appeared in one of his 1912 articles) but does not appear in the 1926 English version. As we’ll see below, the decision to exclude some images from the 1926 version might have served to make it seem more scholarly than its predecessor. Image: Heilmann (1916).

Fighting, gliding and climbing proavians, depicted in a cluttered woodland environment, by Gerhard Heilmann. This is typical of Heilmann’s many excellent scenes. He used posed plaster models of the creatures before drawing. This illustration was used in the 1916 Danish version of the book (it first appeared in one of his 1912 articles) but does not appear in the 1926 English version. As we’ll see below, the decision to exclude some images from the 1926 version might have served to make it seem more scholarly than its predecessor. Image: Heilmann (1916).

Heilmann was an artist and graphic designer by profession but his interest in bird evolution and anatomy was such that he published a series of articles on the subject for the Danish Ornithological Society between 1912 and 1916 (Nieuwland 2004, Ries 2007).

Self-portrait of Gerhard Heilmann, produced in 1902. Heilmann was an argumentative man often in conflict with “various forms of authority”, and at loggerheads with his own family   (Nieuwland 2004)  . This illustration was included in     Nieuwland (2004)  .

Self-portrait of Gerhard Heilmann, produced in 1902. Heilmann was an argumentative man often in conflict with “various forms of authority”, and at loggerheads with his own family (Nieuwland 2004). This illustration was included in Nieuwland (2004).

Compiled, these articles described Heilmann’s take on the diversity of ancient fossil birds, embryological development in birds and reptiles, and bird anatomy, with the last in the series explaining what these data meant for the appearance and lifestyle of his Proavis, an imagined animal older and anatomically more archaic than Archaeopteryx. These articles were published together in a Danish-language book Vor Nuvaerende Viden om Fuglenes Afstamning (Heilmann 1916), the title meaning Our Current Knowledge of the Descent of Birds.

Heilmann’s The Origin of Birds. This body of work was sufficiently impressive that – while Heilmann was mostly shunned in Denmark – it was well received elsewhere, and he was encouraged by scientists in Germany in particular to translate it into English. The result was the 1926 The Origin of Birds (republished in the USA in 1927), a scholarly and well-illustrated book that was well received internationally and almost immediately deemed the most authoritative work on bird origins. The 1926 volume is quite different from that of 1916 in the illustrations it includes (as we’ll see below), but also in how scientific and speculative the text is, some more fanciful sections on palaeobiology being absent from the English version.

It has been said several times that Heilmann’s work was deemed so impressive that, rather than inspire new work on the subject, it effectively bought research on bird origins to a halt. Most relevant workers (albeit not all) now considered the question of bird origins to be resolved. As noted by historian Ilja Nieuwland, things might not have gone this way had the English-speaking readers of Heilmann’s book known that he was an amateur scientist and an artist by trade (Palm 1997, Nieuwland 2004).

Heilmann’s draftmanship was superb. This illustration (fig. 140 from   Heilmann 1926  ) shows thigh feathering on the chicks of various birds, with 8 being a gliding gecko. “The animal with which we may best compare the bird-ancestor, is the Fringed gecko” [sic] (p. 197). Image:   Heilmann (1926)  .

Heilmann’s draftmanship was superb. This illustration (fig. 140 from Heilmann 1926) shows thigh feathering on the chicks of various birds, with 8 being a gliding gecko. “The animal with which we may best compare the bird-ancestor, is the Fringed gecko” [sic] (p. 197). Image: Heilmann (1926).

Heilmann thought that bird ancestors were likely quadrupedal gliders, similar superficially to living gliding geckos (Heilmann 1926, p. 197). This statement is somewhat odd in view of his endorsement elsewhere of a more cursorial view of these animals, but it seems that he imagined them adopting very different poses when on the ground versus when climbing. Heilmann also argued that birds were not dinosaurs – despite his good understanding of the compelling anatomical similarity between theropod dinosaurs and birds – but were instead the descendants of a group closely related to – and supposedly ancestral to – dinosaurs, termed ‘pseudosuchians’*. This was because of his adherence to ‘Dollo’s Law’ – the (erroneous) idea that a lost anatomical structure cannot be regained – and his mistaken contention that dinosaurs lacked clavicles. If dinosaurs couldn’t be ancestral to birds, the true ancestors must have been another group, hence Heilmann’s use of the more archaic ‘pseudosuchians’, already mooted as possible bird ancestors by Robert Broom in his 1913 description of the South African Euparkeria.

* The term pseudosuchian is today applied to the archosaur lineage that includes crocodylians and all of their extinct relatives. Given the historical baggage that comes with the term, I personally don’t think that this is a good idea at all and would prefer it if another name were used for the lineage concerned (like Crurotarsi)… but that’s an issue for another time.

Euparkeria capensis  has long been intimated as a sort of bird ancestor by those looking for such creatures outside of theropod dinosaurs. Not only is this animal only very distantly related to birds, it’s not even part of the crown-archosaur clade. Image: Taenadoman, CC BY-SA 3.0 ( original here ).

Euparkeria capensis has long been intimated as a sort of bird ancestor by those looking for such creatures outside of theropod dinosaurs. Not only is this animal only very distantly related to birds, it’s not even part of the crown-archosaur clade. Image: Taenadoman, CC BY-SA 3.0 (original here).

By combining features common to Archaeopteryx as well as to the ‘pseudosuchians’ Aetosaurus, Euparkeria, Ornithosuchus and Saltoposuchus, Heilmann (1916, 1926) described how he invented a creature that looked somewhat like a theropod but had a more archaic skull and foot, retained a fourth metacarpal, and had a smaller pelvis with far shorter pubic and ischial bones.

Heilmann’s skeletal reconstruction of Proavis, as depicted in the 1916 Danish version of the book… but not in the 1926 English version. Note the lack of long feathers on the hindlimbs. Incidentally, note also that Heilmann was partly responsible for encouraging the belief that forelimb feathers did not cover the hands during the earliest stages of bird evolution. Image: Heilmann (1916).

Heilmann’s skeletal reconstruction of Proavis, as depicted in the 1916 Danish version of the book… but not in the 1926 English version. Note the lack of long feathers on the hindlimbs. Incidentally, note also that Heilmann was partly responsible for encouraging the belief that forelimb feathers did not cover the hands during the earliest stages of bird evolution. Image: Heilmann (1916).

Heilmann, Thompson and D’Arcy Thompson grids. It’s a matter of some interest that Heilmann didn’t just guess what the imagined morphology of the proavian would be, as might be assumed given the English edition of his book. Instead, he used a technique whereby the relevant anatomical regions were mapped on to a grid and then distorted to mimic the evolutionary process. This grid-based deformation process was pioneered by D’Arcy Wentworth Thompson (1860-1948) and was explained most thoroughly in his 1917 book On Growth and Form (Thompson 1917). It was actually put forward beforehand in 1915 (Thompson 1915).

Thompson applied his Cartesian grid technique to many animal lineages, his aim being to show that mathematically predictable transformation in one or more anatomical regions could result in profound anatomical change. Here’s one of my favourite examples: how you can derive a molid sunfish ( Mola  was known to Thompson as  Orthagoriscus ) from a porcupinefish. Image: Thompson (1917).

Thompson applied his Cartesian grid technique to many animal lineages, his aim being to show that mathematically predictable transformation in one or more anatomical regions could result in profound anatomical change. Here’s one of my favourite examples: how you can derive a molid sunfish (Mola was known to Thompson as Orthagoriscus) from a porcupinefish. Image: Thompson (1917).

By placing grids on top of diagrams of related animals (say, the skull of a human and a chimp), Thompson showed how the skewing of the grid in a certain direction “would result in the sort of changes that would allow the emergence of a new species” (Naish 2017, p. 116). These grids are generally termed D’Arcy Thompson transformation grids, Cartesian transformations or Cartesian grids, and their use was quite popular in the evolutionary literature of the early 20th century. Heilmann was inspired to use the technique after Thompson wrote to him about bird evolution in 1915, and the two wrote to each other on many occasions about the technique and its application to ideas on the evolution of horses, hominids and birds. They also discussed how the resulting illustrations could be used in Thompson’s On Growth and Form (Ries 2007).

Heilmann included these Cartesian transformations in the 1916 version of his book, but they weren’t included in the 1926 English version. His take on Proavis was not, therefore, simple guesswork. Image: this montage is from Witmer (1991) but the originals are from Heilmann (1916).

Heilmann included these Cartesian transformations in the 1916 version of his book, but they weren’t included in the 1926 English version. His take on Proavis was not, therefore, simple guesswork. Image: this montage is from Witmer (1991) but the originals are from Heilmann (1916).

Heilmann’s commitment to the technique is demonstrated by the fact that the 1916 Danish edition of his book includes grids that depict the inferred evolution of the avian skull, forelimb and pelvis (I don’t own a copy of the Danish edition, but the diagrams are included in both Witmer (1991) and Ries (2007)). With Ornithosuchus and Euparkeria as ‘starting points’ and Archaeopteryx and modern birds as ‘end points’*, Heilmann used the grids to create relevant intermediates, the results allowing him to generate, piecemeal, his Proavis (Heilmann 1916). A lateral view of the hypothetical animal – showing it in a neural, non-dynamic walking pose – was included in the Danish edition (Heilmann 1916, Ries 2007).

* It should be noted that Heilmann’s Archaeopteryx was not accurate, some of its details (in the skull especially) being semi-hypothetical and informed by Heilmann’s interpretation of ‘pseudosuchians’ like Aetosaurus.

The more dynamic, climbing version of Heilmann’s Proavis skeletal reconstruction, the only version of the reconstruction included in the 1926 English version of his book. Image:   Heilmann (1926)  .

The more dynamic, climbing version of Heilmann’s Proavis skeletal reconstruction, the only version of the reconstruction included in the 1926 English version of his book. Image: Heilmann (1926).

Remarkably, none of this was included in the English edition at all, this creating the impression that Heilmann simply invented the proavian via intuition and guesswork. He didn’t: it was generated via a testable, repeatable method (albeit using flawed anatomical data). Furthermore, the fact that Heilmann’s skeletal reconstruction of Proavis included in the English edition (Heilmann 1926) is the more dynamic climbing version of the creature gives it a less scientific, more speculative, artistic air than the walking version of 1916.

Heilmann, Beebe and Tetrapteryx. When it came to feathering and other aspects of the integument, Heilmann noted his thoughts on another Proavis-type animal, this time the one invented by Charles William Beebe in 1915. Beebe (1877-1962) was an interesting person. He was a naturalist, ecologist, ornithologist, marine biologist, author and explorer, among other things. Today, he might be best known for the deep-water observations he made while in a bathysphere off the coast of Nonsuch Island off Bermuda during the 1930s.

William Beebe’s hypothetical tetrapteryx creature. A pre- Archaepteryx  glider. You’ll note that Beebe was a pretty good artist [UPDATE: this is an error. The illustrations in this publication were by Dwight Franklin, not by Beebe!]. Image: Beebe (1915).

William Beebe’s hypothetical tetrapteryx creature. A pre-Archaepteryx glider. You’ll note that Beebe was a pretty good artist [UPDATE: this is an error. The illustrations in this publication were by Dwight Franklin, not by Beebe!]. Image: Beebe (1915).

When it came to bird origins, Beebe thought that birds originated via a ‘tetrapteryx’ phase where large feathers on the proximal hindlimb were present and functioned in slowing descent during leaping and gliding (Beebe 1915). Both the fossils of Archaeopteryx and the nestlings of living bird species demonstrated the antiquity of these ‘pelvic wings’, Beebe (1915) said. Heilmann (1926) disagreed. His own examination of bird nestlings pertaining to bird lineages across the family tree resulted in his “complete disappointment, for what I found was wholly negative; there was not in any of them the slightest trace of a “pelvic wing”” (p. 194).

As for Archaeopteryx, Heilmann (1926) argued that the hindlimbs were arranged such that they could never have had an aerodynamic function, nor was there evidence for long hindlimb feathering of the sort consistent with ‘pelvic wings’. Ergo, Heilmann illustrated Archaeopteryx without long hindlimb feathering, and he didn’t think that Proavis had them either: his skeletal reconstructions (both the 1916 walking version and 1926 climbing one) lack them (the 1926 climber has short hindlimb feathers, similar in length to those elsewhere on the body) and the gliding proavians included in his various drawings clearly lack them as well. His most famous proavian figure – the semi-erect one standing in a conifer tree (shown below) – does have long feathers on the rear edge of the thigh, though they certainly aren’t as long as they should be as per Beebe’s tetrapteryx model.

The iconic tree-dwelling proavian. The iconic tree-dwelling proavian I have in mind here is one of Heilmann’s best-known and most frequently reproduced illustrations. It has what appears to be great symbolic significance if you’ve read the English version of The Origin of Birds since it’s the very last image in the entire book (Heilmann 1926, p. 199), thus appearing as a visual metaphor for an imagined phrase: “Finally, I can reconstruct the real ancestor of birds, and it looked like this”.

Heilmann’s most iconic version of his Proavis. Note the long feathers on the back of the thigh and how they’re invisible and thus not deemed important in the gliding individual in the background. Heilmann said that the long feathers close to the base of the tail would produce the required lift in this region. Image:   Heilmann (1926)  .

Heilmann’s most iconic version of his Proavis. Note the long feathers on the back of the thigh and how they’re invisible and thus not deemed important in the gliding individual in the background. Heilmann said that the long feathers close to the base of the tail would produce the required lift in this region. Image: Heilmann (1926).

I was surprised to learn that Beebe produced what looks like his own version of this illustration, and inserted it as a plate in a book published in 1915. However, the image itself is not dated 1915 and does not definitely pre-date Heilmann’s illustration. My suspicion – and that of Paul Stewart, who kindly brought my attention to the image – is that Beebe copied Heilmann’s proavian for his own personal use (the illustration was not intended for publication*). This in itself is interesting since it could mean that Beebe personally endorsed Heilmann’s view of Proavis and regarded it as a ‘descendant’ of his tetrapteryx creature… which it wasn’t, given that Heilmann rejected Beebe’s idea, as we saw earlier. Again, my thanks to Paul Stewart for sharing this image with me and permitting its use here.

* Thanks to Jonathan Kane, I’ve learnt that it was previously published in Tim Berra’s 1977 William Beebe: An Annotated Bibliography.

William Beebe seems to have produced his own take on Heilmann’s iconic proavian scene. We’re presuming that it was produced some time after Heilmann’s illustration was, but this hasn’t yet been fully confirmed. The illustration appears here courtesy of Paul Stewart.

William Beebe seems to have produced his own take on Heilmann’s iconic proavian scene. We’re presuming that it was produced some time after Heilmann’s illustration was, but this hasn’t yet been fully confirmed. The illustration appears here courtesy of Paul Stewart.

Today, the significance of Beebe’s tetrapteryx idea is a bit uncertain and, frankly, depends on who you ask. Archaeopteryx probably did have long feathers on the hindlimbs (Longrich 2006), and they perhaps had an aerodynamic role if this animal indulged in aerial locomotion (which it likely did). Furthermore, the discovery of prominent ‘hindlimb wings’ in other feathered dinosaurs (most notably the Chinese dromaeosaur Microraptor) has led some experts to note the predictive power of Beebe’s suggestion (Kane et al. 2016) and to favour the presence and importance of hindlimb feathers in the earliest birds (Longrich 2006, Chatterjee & Templin 2007, Zheng et al. 2013, Xu et al. 2014). However, it’s not at all certain that large, aerodynamically ‘functional’ hindlimb feathers were present consistently across the lineages concerned or that they had the role that Beebe imagined (O’Connor & Chang 2015).

Recent work shows that  Archaeopteryx  really did have long feathers on the proximal portions of its hindlimbs (these reconstructions are by Longrich (2006)), though they weren’t as long as the feathers Beebe imagined for his tetrapteryx stage creature. Image: Longrich (2006).

Recent work shows that Archaeopteryx really did have long feathers on the proximal portions of its hindlimbs (these reconstructions are by Longrich (2006)), though they weren’t as long as the feathers Beebe imagined for his tetrapteryx stage creature. Image: Longrich (2006).

As always, there’s more to say and this article is already longer than intended. I also wanted to talk about Pycraft’s initial concept of the Proavis, and also Burian’s depiction, since he didn’t simply copy Heilmann’s illustrations but actually produced yet another hypothetical creature. These things will have to wait to another time. Also worthy of further discussion is the point made earlier about Heilmann’s influence happening despite his status as a ‘mere’ artist, a theme that has parallels elsewhere in the history of vertebrate palaeontology and is relevant to my writings on meme perpetuation, the scientific acceptance of feathering in non-bird dinosaurs and much else besides. I aim to explore these topics (and others) in future articles.

Before Heilmann, there was Pycraft’s proavian of 1906. We’ll have to discuss this creature and its backstory another time. Image:   Pycraft (1910)  .

Before Heilmann, there was Pycraft’s proavian of 1906. We’ll have to discuss this creature and its backstory another time. Image: Pycraft (1910).

On which note, please consider supporting this blog at patreon if you don’t already do so. The more support I receive, the more time I can spend generating new content. Thanks so much.

For previous TetZoo articles relevant to the issues covered here, see…

Refs - -

Beebe, C. W. 1915. A tetrapteryx phase in the ancestry of birds. Zoologica 2, 38-52.

Chatterjee, S. & Templin, R. J. 2007. Biplane wing planform and flight performance of the feathered dinosaur Microraptor gui. Proceedings of the National Academy of Sciences 104, 1576-1580.

Heilmann, G. 1916. Vor Nuvaerende Viden om Fuglenes Afstamning. Unknown publisher, Copenhagen.

Heilmann, G. 1926. The Origin of Birds. Witherby, London.

Kane, J., Willoughby, E. & Keesey, T. M. 2016. God’s Word or Human Reason? An Inside Perspective on Creationism. Inkwater Press.

Longrich, N. 2006. Structure and function of hindlimb feathers in Archaeopteryx lithographica. Paleobiology 32, 417-431.

Naish, D. 2017. Evolution in Minutes. Quercus, London.

Nieuwland, I. J. J. 2004. Gerhard Heilmann and the artist’s eye in science, 1912-1927. www.PalArch.nl., vertebrate palaeontology 3, 2.

O’Connor, J. & Chang, H. 2015. Hindlimb feathers in paravians: primarily “wings” or ornaments? Biology Bulletin 42, 616-621.

Palm, S. 1997. The Origin of Flapping Flight in Birds. Svend Plam, Ballerop.

Pycraft, W. P. 1910. A History of Birds. Methuen & Co, London.

Ries, C. J. 2007. Creating the Proavis: bird origins in the art and science of Gerhard Heilmann 1913-1926. Archives of Natural History 34, 1-19.

Thompson, D. W. 1915. Morphology and mathematics. Transactions of the Royal Society of Edinburgh 50, 857-895.

Thompson, D. W. 1917. On Growth and Form. Cambridge University Press, Cambridge.

Witmer, L. M. 1991. Perspectives on avian origins. In Schultze, H.-P. & Trueb, L. (eds) Origins of the Higher Groups of Tetrapods: Controversy and Consensus. Cornel University Press (Ithaca, London), pp. 427-466.

Xu, X., Zhou, Z., Dudley, R., Mackem, S., Chuong, C.-M., Erickson, G. M. & Varricchio, D. J. 2014. An integrative approach to understanding bird origins. Science 346 (6215), 1253293.

Zheng, X., Zhou, Z., Wang, X., Zhang, F., Zhang, X., Wang, Y., Wei, G., Wang, S. & Xu, X. 2013. Hind wings in basal birds and the evolution of leg feathers. Science 339, 1309-1312.