Many thanks to Mark Evans of the New Walk Museum, Leicester, for this photograph of yet another camel skeleton, this one from the MNHN in Paris, France:


Head and neck of Bactrian camel (Camelus bactrianus) in right lateral view. Photograph by Mark Evans.


This is especially interesting because it’s our first Bactrian camel — the Cambridge Camel and the Oxford camel are both dromedaries.  I’d wondered whether one species might have a better articulating cervical skeleton than the other, but it seems there is little or nothing to choose between them.

Also of note in this photo is the juvenile camel giraffe in the background, which has thoughtfully been mounted with its nuchal ligament in place.  It’s interesting to see how this ligament has branches that insert separately on the neural spines of all seven cervical vertebrae.  Note, too, that the intervertebral cartilage seems to have been left in place.  This would be good to see in the flesh …  sigh … another reason to revisit Paris, the most hostile city in the world.

Getting back to the adult dromedary in the foreground, here’s a zoom into the joint between the second and third cervicals, with the background smoothed out between them so that you can more easily see the gape between the centra:


Articulated cervical vertebrae 2 and 3 of Bactrian camel (Camelus bactrianus) in right lateral view. Photograph by Mark Evans.


And remember that, once more, the posture adopted for the skeletal mount is much less strongly flexed than the habitual posture in life.  And other postures also adopted in life are more extreme still:


Male dromedary camels (Camelus dromedarius) in rut, with extended necks. Photograph by Gordon Grigg.


This photograph, kindly provided by Gordon Grigg of the University of Queensland, shows typical rutting behaviour of dromedary camels, which he observed closely for his recent paper on the role of strategic hypothermia in reproductive success (Grigg et al. 2009).  The more distant of the two camels is in a truly ridiculous pose.  And it’s ever siller when we bear in mind that necks lie: knowing what we do about the trajectory of the cervical vertebral column within the fleshy neck of tetrapods, it seems likely that the cervical skeletons of these animals were posed something like this:

If you compare these postures with the one that I photoshopped in the Cambridge Camel post, you’ll see that these are more extreme.  I ought to ‘shop the Cambridge camel vertebrae into this pose some time and see just how dumb it looks.

But of course, this may not be as extreme as camel neck poses get.  A few times in recent articles and comments, we’ve alluded to the camel-neck illustration from Kent Stevens’s 2005 talk to the German Research Group on Sauropod Biology at the Sauriermuseum in Aathal, Switzerland.  For those who don’t want to download the complete set of slides, here is that illustration:

We don’t know the provenance of this picture — or, given the low resolution — even whether it’s a photograph or a drawing.  But if it’s real, it’s … stunning.

Anyone know where it’s from?  [Update: see Jerry Harris’s comment below.]


  • Grigg, Gordon, Lyn Beard, Birgit Dörges, Jürgen Heucke, Jocelyn Coventry, Alex Coppock and Simon Blomberg.  2009.  Strategic (adaptive) hypothermia in bull dromedary camels during rut; could it increase reproductive success?  Biology Letters 5:853-856, first published online 15 July 2009.  doi: 10.1098/rsbl.2009.0450

Welcome to post four of what seems to be turning out to be Camel Week here on SV-POW!.  As it happens, I spent last Friday and Saturday in Oxford, for a meeting of the Tolkien Society, and I had three hours or so to spend in the wonderful Oxford University Natural History Museum.

In a completely ideal world, I would have been able to play with a sequence of camel cervicals; but very short notice, the collection manager’s unavailability and the off-site location of many specimens conspired to prevent this.  And now that the museum’s online specimen-catalogue search is back up, I see that they don’t in fact have a sequence of camel cervicals.  But they do have a mounted camel, and I was able to take a good look at its neck.

And now, so can you:

Head and neck of dromedary camel (Camelus dromedarius), OUMNH 17427, in left lateral view.

At first glance it doesn’t look as silly as the Cambridge Camel — it doesn’t have those gaping spaces between the centra of the neck.  But on second glance, you can see that the reason for this is that the mount has fat wodges of fake articular cartilage wedged between the centra.  Take a closer look:

Cervical vertebrae 2-4 of dromedary camel (Camelus dromedarius), OUMNH 17427, in left lateral view, with thick false-cartilage spacers between the centra.

As you can see, those things are thick.  It’s not easy to measure them, or the vertebrae, in a fairly tall mount that you’re not meant to touch, and which is up on a pedestal, behind a rope that you’re not allowed to cross.  But because I love you guys, I did the best I could with my foldable Ikea paper tape-measure, and here are the figures I came up with (omitting C1, which I couldn’t reach):

  • C2: 18 cm long, with 3 cm of “cartilage” behind it
  • C3: 17 cm / 3 cm
  • C4: 17 cm / 2 cm
  • C5: 17 cm / 2 cm
  • C6: 14 cm / 1.5 cm
  • C7: 11 cm / 1 cm

Adding these up gives us 12.5 cm of cartilage for the 94 cm of neck — adding 13% on to the length of the actual vertebrae.  As a side-note, IF sauropods had a similar ratio of cartilage, then the 15 m neck of Supersaurus would be more like 17 m.  But that is a very big supposition, and not one I am going to try to defend … not only because sauropods ain’t mammals but also because the idea of a solid six inches of cartilage between adjacent vertebrae is a little bit scary.

[Oddly enough, we’ve featured the Oxford Camel before on SV-POW! — more than two years ago, in fact.  It’s in the background of the second photo in our old SV-POW! on tour post, standing behind Matt, Darren and me, and showing off its false cartilage pretty darned clearly if you click through to the full-sized version.  But we paid no attention to its barely-longer-than-a-meter neck, because we were young and stupid back then.]

Cervical vertebra 5 of dromedary camel (Camelus dromedarius), OUMNH 17427, in left ventrolateral view, with thick false-cartilage spacers before and after its centrum.

What does this tell us?

At this stage, it’s really just another chunk of ignorance to chuck onto the Big Old Heap Of Ignorance (hereafter, the BOHOI).  But the point is that at least now we know we’re ignorant.  That’s progress, isn’t it?  Isn’t it?  Please tell me it is.  Whatever else it may tell us, the Oxford Camel is more evidence that the characteristic posture and range of motion of extant animals are not constrained by, and do not closely resemble, poses determined purely from osteology.  It doesn’t, yet, get us much further than that, though — all it does it make us more aware of how much more work there is to be done.  Just quantifying the error would help.  We demand rigidly defined areas of doubt and uncertainty.

I leave you with this picture, a larger version of one that Darren once used on Tetrapod Zoology, and which he sent me a couple of days ago:

That is some serious flexibility: as well as bending its neck to the right, this deer has twisted it through 180 degrees — not something that it’s obviously capable of from looking at its skeleton.

How is it done?  We don’t know.  But we aim to find out.

Coming soon …

… Sauropod vertebrae!  Yes, really!

Maybe all the camels are wrong

September 27, 2010

Suddenly it’s camel season here at SV-POW! In the last post, Mike was having some doubts about how far back camels could get their heads. That got me curious, so here are the results of 45 minutes worth of Google Image Search.

This live baby camel (source) has its neck extended about as far as the presumably dead juvenile camel from the last post, so that pose is not just mechanically possible, but also achievable in life. Admittedly, this is a baby, so it might have a bit more flexibility than adults. But I doubt if it is really pushing things here, since I’ve seen adults get into more extreme poses than this.

This dromedary Bactrian camel [thanks to John Scanlon for the correction] (source) has its head back pretty close to the hump, but it’s hard to tell with all the hair. People trying to work out the normal neck postures of the members of KISS probably run into the same problem.

I think we should have a caption contest for this one (source). Not terribly informative from a figuring-out-what-the-bones-are-doing point of view, but I like it.

A kneeling camel, from Werner Herzog’s Even Dwarfs Started Small, described as “the only film whose all-dwarf cast gleefully commit acts of arson and vandalism, throw themselves at or on top of moving vehicles, flick through pornographic magazines, stage a mock religious ceremony with a crucified monkey and laugh themselves hoarse at a camel as it takes a dump” (source, for both photo and quote). What’s cool here is that the camel’s back is angled down at the front, but its neck and head posture are unchanged.

Same thing, this time in color, sans dwarf (source).

Now, none of these camels are maxing out because they don’t have their heads laying back on their humps, as shown in slide 9 of this talk by Kent Stevens (thanks to Steve O’C for the link, in a comment on the last post). Which prompted Mike to write:

Yes, I’ve seen the picture you allude to, with the dorsal surface of the camel’s head in contact with the hump. (In fact I am pretty sure that either John or Kent showed that picture at SVPCA.) I initially planned to photoshop the mounted camel’s cervicals into that position, but it just seemed too ludicrous to believe in, and I’ve come to suspect that this picture might itself be the work of photoshoppers.

Look kids, the argument from personal incredulity!

It’s not Photoshop. There are more photos like this, they’re just not easily available online (believe me, I tried). The one shown by Jeff Wilson in his Jobaria talk at SVP ’99 was of a whole live camel on location in the desert, not the same posed camel you can see in Kent’s talk. Also, I’ve seen camels do this at the circus.

And why is it so hard to believe that camels can get their necks back that far? As Mike pointed out in another comment, the anterior verts aren’t extended much at all in his Cambridge camel. If they have anything like the flexibility of the posterior cervicals, getting the head back against the hump ought to be a cinch.

I am starting to think that camels might be the most interesting mammals out there. The neck of the giraffe certainly looks like it is suspended from the withers, whereas camel necks aren’t connected by any straight-line ligament from the back of the head to what pathetic withers they have (meaning that they do have a nuchal ligament, but it can’t be working like a suspension bridge cable inside that curvy neck), and must be held in those ridiculous curvy poses by continuous muscular effort. But when you look at the cervicals, there are no neural spines at all through the middle of the neck! Not to mention the very flat zygs that look like they shouldn’t allow the poses in the first place. It’s like they’re defying us to make any sense of them.

Clearly, what we need to do is visit museums with complete but disarticulated camel necks that we can put in sandboxes and pose, like I did for the chicken and the infamous rabbit way back when. It’s no good taking photos of mounted skeletons and declaring that they’re in ONP. Zyg-by-zyg scrutiny often reveals that one or more joints is not in ONP (usually more, if the animal is mounted in anything like a normal lifelike pose), and the spaces between the centra are often filled with some weird goop that is supposed to look like intervertebral cartilage, or just left open as in the Cambridge camel. Here is another shot from that talk of Kent’s, of a camel neck supposedly in ONP, illustrating both problems.

In the same talk, Kent wrote that the zygs of the camel do allow the head-to-hump posture. That’s backed up by some Photoshopped images of the same mounted camel with the goop in between the verts. I have no doubt that the craniocervical system of the camel allows the head to touch the hump, because I’ve seen it with my own eyes. What I’d like to know is whether you could put a disarticulated camel neck in a sandbox and achieve the same pose without violently disarticulating the cervicals. The photos of the Cambridge camel suggest that either the zygs are going a lot farther past each other than is commonly assumed, or the intervertebral cartilage is allowing more separation of the centra.

In Kent’s talk at SVPCA he cited a pers. comm. from Kent Sanders and me at SVP ’98, that in playing with ostrich necks we could not get the zygs to completely disarticulate, and that the bone would break before that would happen. That’s true, that is what we found. But the really important part of what we found is that the zygs don’t stay parallel to one another. That is, in flexing and extending the neck, the cervical zygs don’t just slide past each other in the same plane, they can also hinge apart like the covers of a book. You’ll recall that the assumption in Stevens & Parrish (1999) was that the zygs maintained 50% overlap, but subsequent work (including work by them) has shown that much smaller overlaps are possible. My work with Kent Sanders on ostrich necks suggests that the problems of determining ROMs from bones are even worse, because the zygs can get to 20-25% overlap and then hinge open, so that only the very edge of one zyg is still in contact with the other. At that point it is meaningless to even talk about overlap. How you constrain that in your model, I have no idea.

Finally, I have a memory of Greg Paul saying years ago, possibly on the DML, that if you fed a camel into DinoMorph, it would crash the program. If anyone can find that quote, I’d be grateful. To my knowledge, that assertion has never been tested, although I think it would be an informative exercise for all parties and I would be most interested in the results.

Bonus Sauropod Image

Because it was a long dry summer (Gilmore 1932:pl. 6):


  • Gilmore, C. W. 1932. On a newly mounted skeleton of Diplodocus in the United States National Museum. Proceedings of the United States National Museum 81, 1-21.
  • Stevens, Kent A., and J. Michael Parrish. 1999. Neck posture and feeding habits of two Jurassic sauropod dinosaurs. Science 284: 798-800.

Since I posted my photograph of the Cambridge University Zoology Museum’s dromedary camel in the last entry, I haven’t been able to get it out of my mind.  Here it is again, this time with the background removed:

You’ll remember from last time that the thing that struck me most powerfully about it was the huge disarticulations between the centra of C3, C4 and C5.  [Stevens and Parrish (2005:fig. 10.1A) illustrated the articulated cervical column of a dromedary camel Camelus dromedarius in osteological neutral pose, and it comfortably approximates life posture; but its vertebrae are very different from those of this specimen.  I don’t know what to make of this.  Are there dromedary subspecies?  If so, they are very different from each other; if not, then the individual variation is pretty amazing.]

The Cambridge mount made me wonder how the neck of that specimen would look if we moved it down into neutral pose — that is, keeping the zygapophyses maximally overlapped as they are in the mount, but bringing the centra together at the same time.  I tried it in GIMP (a free equivalent to the better-known PhotoShop), and here is the result:

Let’s be clear that photoshopping vertebrae is an inexact science at best: I am working here from a single photograph taken carelessly as one among a hundred taken opportunistically in a museum too awesome not to photograph; I can see the vertebrae only from one angle; judging the maximal zygapophyseal overlap is error-prone.

Still, even taking all of these factors into account, I found this pose striking.  It left me very much wanting to find a published osteology of the camel with better multi-view figures of the cervical vertebrae.  Sadly, it seems like there isn’t anything like that (though if you know better, PLEASE say so in the comments!)  But my search led me inevitably to tetrapod savant Darren Naish, and he pointed me to Maziersky (2010), a book review which includes the following photo:

Camel with raised neck, from Maziersky (2010:fig. 2)

Judging by the odd way the camel is propped up on a table, this is a dead animal being posed rather than a live one adopting a posture voluntarily, but it does appear that this is at least a pose that the mechanics of the animal allow.  And that got me thinking about how the vertebrae must be arranged to allow this.  Here’s the best I’ve been able to come up with:

In comparison with the mounted skeleton’s pose, this re-articulates cervicals 3 and 4; but 4 and 5 remain horribly disarticulated, and the 5-and-6 and 6-and-7 pairs are now also in this state.

(A reminder is due again here that what I am doing is an approximate and error-prone process.  No doubt I got the maximal possible zygapophyseal disarticulation wrong in several places, for example.  But even allowing for that, I find this pretty amazing.)

If you’re wondering why the two earlier images had so much blank space at the top and this one has so much to the right, it’s because I made them all the same size and shape.  This means that if you open all three images in different browser tabs, then tab between them, you should see the neck neatly moving between the three different poses.  For those of you too lazy or technophobic to do that, here is a superposition:

Habitual posture (i.e. when the animal is not eating or drinking or otherwise doing anything in particular with its head) is somewhere above the mounted pose, but less extended than the raised pose shown by Maziersky.

What does all this tell us?

Nothing very encouraging, I’m afraid.   Even allowing for the vagaries of photoshopping images of museum mounts, it’s apparent that something very weird is going on in this camel’s neck, such that even a pose well below the habitual one requires extensive vertebral disarticulation.  Assuming that, like me, you don’t believe the vertebrae really are disarticulating in life, we can only conclude that it is useless to try to reach conclusions about neck posture based on osteology alone.  We need to understand the soft-tissue systems — especially the articular cartilage — as well.

Stevens and Parrish (1999:798) stated that “in vivo, muscles, ligaments, and fascia may have further limited movement [i.e. beyond the restrictions imposed by maintaining zygapophyseal overlap]; thus, the digital manipulations reported here represent a ‘best case’ scenario for neck mobility.”  Although this seems intuitively appealing, evidence including but not limited to the Cambridge camel shows that the opposite is actually the case: in at least some taxa, and maybe all, soft tissue enables necks to be more flexible, not less, than the bones alone suggest.

Folks, we’re flying blind.  Until we start to understand the soft tissues in the necks of extant critters — especially the intervertebral cartilage, but I bet that’s not the whole story — we really have no idea how to interpret the bones.

Come on, neontologists!  Teach us about intervertebral cartilage!


Special bonus archosaur-rich artwork

Check out John Conway’s obscenely brilliant infinite-zoom Jehol video.  (Well, a lot of people have been calling it infinite zoom, but it’s clearly finite.  Still, it’s at least Very Big Zoom.)  A lot of jaws dropped at SVPCA in Cambridge when John was showing this off.  While you’re at it, you might like to read the interview with John at Dave Hone’s Archosaur Musings.  Dave’s interviewed quite a few palaeoartists now, but John has more to say than most of them, and it’s well worth a read.

Special bonus horror story

While I was emailing with Darren about camels, he told me that John Hutchinson had recently acquired a camel at the RVC, and suggested that I ask to be present at the dissection of the neck.  I contacted John only to be told: “Just got the feet; had no time to get the rest, sadly. Notice came at a bad time for my group, as it tends to do. It is now incinerated.”  John also told me at SVPCA of a hippo that was recently incinerated because he couldn’t get to the zoo to collect it within 24 hours.  Graaaggh!  It’s a tragedy the dead animals that go to waste.

I’m just back from SVPCA 2010 (the Symposium of Vertebrate Palaeontology and Comparative Anatomy), and what an amazing meeting it was.  I think it was the best I’ve been to.  That’s partly because I understand more of the talks these days — it’s the first time I’ve ever listened to every single talk, even all the mammal-tooth and fish-skull talks — and I learned something interesting and new from almost every one of them.

But as is so often the case, the best thing about the meeting was, well, meeting.  I met with Matt and Darren for the first time in a year, which is always excellent.  And for the first time, I met horizontal-sauropod-neck advocate Kent Stevens.  Kent was there to present one of two talks on horizontal necks, and UK sauropod jockey John Martin presented the other.  Their talks were part of a block of seven sauropod talks — it would have been eight had Michael Pitman not changed his scheduled sauropod-tail talk to a theropod-tail talk.  Matt and I both made presentations, although Darren wasn’t able to because he didn’t know that he’d be able to come to the meeting until the last moment.

After that block of talks, Matt, Darren and I went off to lunch with Kent and Martin.  Despite the lighthearted attempts of session moderator John Hutchinson to build the session up as a two-way fight, it was all rather peaceful and enjoyable.  After lunch we all went to have our photos taken together in front of the Zoology Museum‘s giraffe skeleton:

Sauropod Neck Posture Working Group, 2010 meeting.  From left to right: Darren Naish, Matt Wedel, John Martin, Mike Taylor, Kent Stevens.

As you can see, we were all very civilised and well behaved.

The Sauropod Neck Posture Working Group carefully considers all points of view in a detached, professional and mature manner.

In all seriousness, it’s no secret that we SV-POW!sketeers are very much advocates of a raised habitual posture, and so that we strongly disagree with Kent and John.  We had a lot of fun talking together, but we didn’t find that they presented any compelling new evidence in their talks.  (You can read the abstracts of their talks, and indeed of mine and Matt’s, in the SVPCA abstracts book.)

The case for horizontal or near-horizontal habitual pose rests on two assumptions.  First, that osteological neutral pose (ONP) was habitually adopted; and second, that we can know what ONP was.  We still feel that both of these assumptions are false.  We can’t know ONP because there is not a single sauropod neck skeleton anywhere in the world consisting of undistorted cervicals — and even if we knew what ONP was, it wouldn’t tell us much about what I am suddenly going to call mechanical neutral pose (MNP)[*], because we don’t know anything about the intervertebral cartilage.  And we know that extant animals do not habitually adopt ONP because we have X-rays that show us how they habitually rest, and we know that they don’t match what you get by articulating bones.

[* either John or Kent made the point that ONP != MNP in his talk.  I think they probably used a different name for MNP, but it eludes me for now.  If anyone can remind me, I will switch to their terminology.]

So, anyway, it was a bit frustrating watching John’s talk, and seeing him show many photographs of live animals and claiming that their necks were in ONP, when we knew perfectly well that they were not — because necks lie.  We fear he may have been tricked by the misleading soft-tissue outlines that mask the postures adopted by the neck skeleton in nearly all tetrapods.  As an example, I give you the hoatzin, which happily was on display at the Zoology Museum as both a stuffed specimen and a skeleton:

Hoatzin (Opisthocomus cristatus), stuffed specimen and skeleton.  Note the extraordinarily long cervical skeleton, almost entirely unreflected in the live animal.

Here’s another photograph from the astounding collection of the Zoology Museum (and some day I really ought to blog about the museum itself).  I took this photograph of the neck of a camel with no specific agenda, but when I looked at it again today, one aspect leapt out at me:

Head and neck of dromedary camel (Camelus dromedarius) UMZC H.14191, in right lateral view, with disarticulated C3/C4 and C4/C5 joints.

Notice how very dramatically the third and fourth cervical central fail to contact, and the fourth and fifth.  How uncomfortable this must be for the poor camel — its neck extended (or “dorsiflexed”) far, far out of ONP, to the point where the vertebrae drastically disarticulate.  And yet we all know perfectly well that habitual pose for camels is much more extended than this, and many of us have seen photos of camels leaning their necks right back so that their heads are upside down, and they can rub the top of their head against their back.  Just imagine what that does to the cervical articulations.

More on this subject another time.  For now, I leave you with more from the Sauropod Neck Posture Working Group summit.

Hey!  That hurt!