Re: Diapsids

[David Marjanovic <david.marjanovic@gmx.at>]

> <<Well, and *Youngina* doesn't have a middle ear.>>
>
> Can you show me a skull without a stapes? I  haven't seen one yet.

The stapes is there, but doesn't function in hearing -- instead it's a 
mechanical part of the skull architecture, a brace between the braincase and 
the quadrate. This is its plesiomorphic function. Turtles and crown-group 
diapsids have freed the the stapes from the quadrate so it can swing freely 
and (now that its shape has changed) conduct airborne sound (other than 
infrasound at least). Crown-group mammals (twice independently, it seems) 
have instead freed the quadrate from its involvement in the skull, so that 
all three middle-ear ossicles can swing freely.

> <<Probably we are talking convergence here, because *Sphenodon* is a
> derived lepidosauromorph and the others are derived archosauromorphs.>>
>
> Not sure that Sphenodon has ever been shown to be derived, except in the
> context of other sphenodontians. Not in regard to lizards.

It's pretty far away from the root of Lepidosaur_omorpha_. That's all I'm 
saying.

> <<Then maybe the emargination migrated laterally. Just guessing.>>
>
> Maybe better to 'look up,' rather than guess.

Please do that -- I don't have time, sorry.

>         And if homologous in diapsids, does the emargination follow the
>
>         same pattern in all forms?
>
> << Doesn't matter -- if all can be derived from the same pattern.>>
>
> If you say the pattern is homologous in all so-called diapsids,

Sorry for the misunderstanding -- I'm just saying it's homologous within the 
crown-group. *Petrolacosaurus* is far outside.

> << As mentioned above, all that matters is the variation around the
> point in the tree that we're discussing. BTW, Ichthyopterygia and their
> apparently close relative *Hupehsuchus* tend to be found inside the
> ingroup lately.
>
> There's an old phrase in paleontology we need to remember: "Is you is,
> or is you isn't?"  'Tend to be found' is so ephemeral and popular and so
> unscientific. Show evidence, David. Gain respect.

Then wait till I've finished my dissertation, please. I'm supposed to begin 
with it this summer. In the meantime simply browse the most recent 
publications (keywords: 2003, 2004, Senter, Müller).

>         But note that
>         the little digger Bipes retains a pair  of large centralia.
>
>
> << Is the lateral centrale one of them? (Anyway -- deep within the
> ingroup, so most likely doesn't matter.)>>
>
> Let's just sweep away the unwanted data.

Excuse me, the fact that whales don't have hair all over their bodies 
doesn't change the optimization of that character at the origin of Mammalia.

X--A
`--+--A
  `--+--A
     `--+--A
        `--+--A
           `--+--A
              `--B

Which character state did organism X have? A or B?

>         You can
>         see this process in action in Homoeosaurus.
>
> <<Are you sure?>>
>
> Two bones become one. Yes. Right at the node too.

Could you direct me toward an illustration? (Sorry for not looking myself --  
no time.)

>         Convergent fusion also takes place in some
>         outgroup sauropterygians, such as Endennasaurus.
>
> <<Probably part of the ingroup.>>
>
> You're waffling. Is you is or is you isn't?

See above.

(In fact I can't remember any analysis that found ichthyosaurs and 
sauropterygians outside the crown-group, but that surely doesn't mean much.)

>         Youngina SAM K7710 also has this hook.
>
> <<Really? *Youngina* is famous for lacking it.>>
>
> I searched high and low for Youngina pedal data. If you have anything
> like you describe, please send it.

In Paris I had the opportunity to read Broom's original papers, for 
example... I don't have them here.

> <<Well, you have so many extant taxa in your matrix that you're bound to
> get all manner of long-branch attraction unless you put, for example,
> toxicoferan autapomorphies into the matrix. As a primitive example -- if
> I'd put an eagle into my tiny Mesozoic bird analysis without adding any
> neornithean autapomorphies, it would most probably come out as an
> avisaurid, next to *Neuquenornis*, instead of clustering with the
> duck.>>
>
> Now there's a challenge. Turn your  probably (a priori decision) into a
> reality by doing exactly that. Then tells us.
> Remember, the term 'probably' belongs in the realm of gambling,
> insurance and weather, none of which are remotely like paleontology
> where we like to say -- is or isn't -- and then take the slings and
> arrows with chest inflated.

My points don't go away by correctly pointing out that I said "probably" 
instead of "with a posterior probability of 89 %". It's your job to defend 
your analysis -- and that works only by incorporating my challenges, running 
it again, and seeing what (if anything) changes.

If you want me to, I will code an owl (of which I happen to have a skeleton 
illustration -- Feduccia 1996) this weekend, put it into my matrix, and tell 
you all what comes out.