Kopidosaurus: no longer an enigmatic iguanian

Scarpetta 2020 bring us a tiny new Eocene lizard,
Kopidosaurus perplexus (YPM VP 8287) known from most of a disarticulated skull still in the matrix, but carefully presented in several views as µCT scans (Fig. 1).

Scarpetta warns his readers,
“Fossil identifications made in a phylogenetic framework are beholden to specific tree hypotheses. Without phylogenetic consensus, the systematic provenance of any given fossil can be volatile. Pleurodonta (Squamata: Iguania) is an ancient and frequently-studied lizard clade for which phylogenetic resolution is notoriously elusive.“

Scarpetta reports,
“I address the effects of three molecular scaffolds on the systematic diagnosis of that fossil. I use two phylogenetic matrices, and both parsimony and Bayesian methods to validate my results, and perform Bayesian hypothesis testing to evaluate support for two alternative hypotheses of the phylogenetic relationships of the new taxon.”

Scarpetta did not provide a reconstruction of the skull. That is remedied here (Fig. 1).

Unfortunately Scarpetta’s three molecular scaffolds are based on genes, so they are completely useless for deep time studies, as documented several times in vertebrates. None of the three genomic studies in Scarpetta 2020 agree with each other. None agree with the LRT (Fig. 2).

Unfortunately Scarpetta’s phylogenetic analyses result in lists of suprageneric taxa, not genera, as in the LRT. Scarpetta reports, “The uncertainty of the relationships of Kopidosaurus is due in part to the mosaic morphology of the fossil and the problematic nature of pleurodontan phylogeny.”

There is no such thing as mosaic evolution. So stop using that excuse.

It doesn’t have to be this complicated. Use the LRT. It’s simple. Just Plug ‘n’ Play.

Figure 1. Kopidosaurus perplexus in situ and µCT scans from Scarpetta 2020. Reconstructions in lateral and palatal views added here.

Scarpetta reports,
“YPM VP 8287 preserves no morphological feature or combination of features that would allow clear referral to any member of Pleurodonta.” And that’s why he shouldn’t be “Pulling a Larry Martin” (relying on key traits that might converge). Instead: drop the new taxon into a comprehensive cladogram, like the LRT (Fig. 2), and let the software nest the enigma.

Definition according to Wikipedia:
“Pleurodonta (from Greek lateral teeth, in reference to the position of the teeth on the jaw) is one of the two subdivisions of Iguania, the other being Acrodonta (teeth on the top [of the jaw]). Pleurodonta includes all families previously split from Iguanidae sensu lato (Corytophanidae, Crotaphytidae, Hoplocercidae, Opluridae, Polychrotidae, etc.), whereas Acrodonta includes Agamidae and Chamaeleonidae.”

The frontal and parietal are incomplete
and the skull is small at <2cm. Am I the first to wonder if this was a juvenile skull? Scarpetta does not bring up the subject. The large orbit relative to skull length supports that hypothesis. Otherwise this could be an adult in the process of phylogenetic miniaturization, common at the genesis of many clades (Fig. 2).

Scarpetta concluded,
“Given the phylogenetic volatility of Kopidosaurus, I refrain from favoring any biogeographic or divergence hypothesis based on the identification of the fossil and advise similar caution for other systematically enigmatic fossils, lizard or otherwise.”

Don’t give up! Use the LRT.

Here 
in the large reptile tree (LRT, 1740+ taxa) Eocene Kopidosaurus nests at the base of a clade of living iguanians (Fig. 2). It is a plesiomorphic taxon, but that doesn’t matter to the LRT. Only a suite of characters is able to nest Kopidosaurus with this level of confidence by minimizing taxon exclusion.

Figure 2. Subset of the LRT focusing on basal Squamata. Here Kopidosaurus nests at the base of a clade of living iguanians including Pristidactylus, Basisliscus and Anolis.

As a reminder,
the LRT is still using just 238 traits, most of which were not used here due to the lack of a premaxilla, vomer and post-crania. Paleontologists still don’t want to accept the fact that the LRT continues to lump and separate with so few multi-state characters. Even those taxa previously tested without resolution, as described by Scarpetta 2020.

---

References
Scarpetta SG 2020. Effects of phylogenetic uncertainty on fossil identification illustrated
by a new and enigmatic Eocene iguanian. Nature.com/scientifcreports 10:15734.
https://doi.org/10.1038/s41598-020-72509-2

Dorsetisaurus: a Mesozoic tegu, not an anguimorph

Known from the Early Cretaceous of Mongolia
and the Late Jurassic of Portugal, Dorsetisaurus purbeckensis (BMNH R.8129, skull width: 1.4cm; Hoffstetter 1967; Fig. 1) was attributed to the clade of glass lizards (Anguimorpha) originally and in two later papers. Evans 2006 nested it between the highly derived legless skink, Amphisbaenia, and the basal gecko (in the LRT), Chometokadmon (which Evans considered an anguimorph).

FIgure 1. Dorsetisaurus bits and pieces restored here and scored nests in the LRT with Tupinambis, the extant tegu.

By contrast
in the large reptile tree (LRT, 1318 taxa) Dorsetisaurus nests with the basal scerloglossan, lacertoid, teiid, Tupinambis (Fig. 2), the extant tegu lizard. Even the slight notch in the ventral maxilla is retained over 120 million years of evolution.

Figure 2. Tupinambis is the extant tegu lizard, a sister to Dorseitsaurus in the LRT.

On a side note:

Gauthier et al. 2012 put together two squamate trees of life, one based on traits, another based on genes. Neither matches the LRT, which includes more fossil taxa.

References
Evans SE, Raia P, Barbera C 2006. The Lower Cretaceous lizard genus Chometokadmon from Italy. Cretaceous Research 27:673-683.
Gauthier, JA, et al. 2012. Assembling the squamate tree of life: Perspectives from the phenotype and the fossil record. Bulletin of the Peabody Museum of Natural History 53.1 (2012): 3-308.
Hoffstetter  R 1967. Coup d’oeil sur les Sauriens (lacertiliens) des couches de Purbeck (Jurassique supérieur d’Angleterre Résumé d’un Mémoire). Colloques Internationaux du Centre National de la Recherche Scientifique 163:349-371.

wiki/Dorsetisaurus
http://fossilworks.org/bridge.pl?a=taxonInfo&taxon_no=38022

Bipedal Cretaceous lizard tracks

These are the oldest lizard tracks in the world…

(if you don’t consider Rotodactylus (Early Triassic) strictly a ‘lizard’ (= squamate). One rotodactylid trackmaker, Cosesaurus, is a tiny lepidosaur).

Figure 1. Bipedal lizard tracks from South Korea in situ. They are tiny.

From the abstract

“Four heteropod lizard trackways discovered in the Hasandong Formation (Aptian-early Albian), South Korea assigned to Sauripes hadongensis, n. ichnogen., n. ichnosp., which represents the oldest lizard tracks in the world. Most tracks are pes tracks that are very small. The pes tracks show “typical” lizard morphology as having curved digit imprints that progressively increase in length from digits I to IV, a smaller digit V that is separated from the other digits by a large interdigital angle. The manus track shows a different morphology from the pes. The predominant pes tracks, the long stride length of pes, narrow trackway width, digitigrade manus and pes prints, and anteriorly oriented long axis of the fourth pedal digit indicate that these trackways were made by lizards running bipedally, suggesting that bipedality was possible early in lizard evolution.”

Actually, the lizard was not running.
Typically in running tracks the prints are very far apart and these tracks are sometimes left toe to right heel.

Figure 2. Original and new tracings of the bipedal lizard tracks from South Korea. PILs are added. Manual digit 4 and 5 appear to have shifted.

 The authors did not venture who made the tracks.
They reported, “based on the palaeobiogeographic distribution of facultative extant families, the lizard that produced S. hadongensis tracks could well have been a member of an extinct family or stem members of Iguania, which was present in the Early Cretaceous.”

Actually the closest match among tested taxa
is with Eichstaettisaurus (Fig. 1), a basal member in the lineage of snakes. And this clade is close to the origin of geckos. ReptileEvolutiion.com and the large reptile tree would have been good resources for the authors to use. Lots of lizard pedes were illustrated and scored there.

Figure 3. Originally imagined  as a generic lizard (below), here Eichstattsaurus matched and scaled to the track size walks upright.

 Based on a phylogenetic analysis of the tracks

the closest match in the LRT is with Eichstaettisaurus, so a slightly larger relative made them. Distinct from the skeletal taxon, the trackmaker had a longer p2.1 than 2.1 and pedal digit 1 was quite short. Otherwise a good match in all other regards.

So why walk bipedally?

It was walking, not running, so escape from predation can be ruled out. Elevating the upper torso and head, like a cobra, can be intimidating to rivals, or just offer a better view over local plant life. This sort of flexibility could have helped them get into the trees and then to move to higher branches.

References

Lee H-J, Lee Y-N, Fiorillo AR &  LÃ J-C 2018. Lizards ran bipedally 110 million years ago. Scientific Reports 8: 2617. doi:10.1038/s41598-018-20809-z

https://www.nature.com/articles/s41598-018-20809-z

My what big eyes you have, Lyriocephalus!

Looking like some sort of medieval fever dream,
meet Lyriocephalus, the hump-nosed lizard (Fig. 1), a cousin to Draco, the gliding lizard. Distinct from Draco, the body of this insectivore is laterally compressed, not laterally extended.

Figure 1. Lyriocephalus in vivo.

Probably the largest eyes
relative to the skull of any tetrapod. Lyriocephalus, is an arboreal jungle lizard with an anterodorsal naris and a small antorbital fenestra. Note the arching postorbital contacting the prefrontal.

Figure 2. Lyriocephalus skull in several views. Note the arching of the postorbital to contact the prefrontal. And did I mention that antorbital fenestra?

Lyriocephalus scutatus (Merrem 1820) is represented by a skeleton at Morphospace.org where you can rotate the skeleton on your monitor. Note the brevity of the tail of this agamid iguanid, There are more in vivo pix here. And a video here.

Figure 3. Lyriocephalus skeleton from Morphobank.org, where you can rotate digitized skeletons.

References
Merrem B 1820. Versuch cines Systems Amphihien Tentamen Systcmatis Amphibiorum. Marburg, Krieger.

wiki/Draco
wiki/Lyriocephalus

Pristidactylus: a Basiliscus sister without a crest

Figure 1. Pristidactlyus torquatus in vivo.

I got interested in the extant lizard, Pristidactylus
(Figs, 1, 2) when Bever and Norell 2017 used it as an outgroup to the clade Rhynchocephalia. The large reptile tree (LRT, 1122 taxa) using phylogenetic analysis falsifies that hypothesis of relationships.

Figure 2. Pristidactylus skull in 5 views. This iguanid lizard nests with the crested basilisk.

Pristidacatylus torquatus (Phillippi 1861, extant, snout-vent length = 6-11cm) is the extant forest lizard. It is related to Basiliscus and feeds on beetles. Image from Digimorph.org.

Figure 3. Basiliscus, the “Jesus” lizard, does not share as many traits as Draco and Chlamydosaurus do, but is related, given the short list of Iguanids currently employed.

Basiliscus basiliscus (Laurenti 1768) is the extant basilisk. It is related to Iguana but has a tall parietal crest. This frilled lizard is able to run bipedally across ponds. Skull image from Digimorph.org and used with permission.

References
Laurenti JN 1768. Specimen Medicum, Exhibens Synopsin Reptilium Emendatum cum
Experimentis Circa Venena et Antidota Reptilium Austriacorum. Wien.
Linnaeus C 1758. Systema naturæ per regna tria naturæ, secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locis. Tomus I. Editio decima, reformata.
Philippi RA and Landeck L 1861. Neue Wirbelthiere von Chile. Archiv für Naturgeschichte 27 (1): 289-301.

wiki/Iguana
wiki/Basiliscus_(genus)

Mid-Cretaceous lizards in amber from Myanmar

A new paper
from Daza et al. (2016) brings us several lizards in amber from the Mid Cretaceous (Fig. 1).

Figure 1. Mid-Cretaceous lizards in amber from Daza et al. 2016. Highlighted specimens are examined here.

I was chiefly interested in
the unidentified ones, JZC Bu267 (Fig. 2) and JZC Bu1803 (Fig. 4).

Figure 2. JZC Bu 267 nests with the pre-snake, Jucarseps in the large reptile tree.

JZC Bu 267
is a tiny, slender, short-legged, long-toed lizard that nests with the Early Cretaceous pre-snake, Jucarseps, (Bolet and Evans 2012, Fig. 3) in the large reptile tree. This nesting is based on relatively few traits as the skull is largely missing while dermis covers large portions of the post-crania. The two are about the same size and overall proportions.

The authors report, “The preservation of  JZC Bu267 is exceptional; it includes the epidermis and soft tissues and even a unique extended tongue tip with a narrow medial projection that does not resemble the form of any described squamate tongues.” According to Wikipedia, reptiles developed forked tongues in several clades independently.

Figure 3. Jucaraseps in situ. This tiny long lizard is in the lineage of terrestrial snakes and nests as a sister to JZC Bu 267,

So this find
extends the range of the Jucarseps clade from SW Europe to SE Asia.

Figure 4 JZC-Bu1803 has a relatively large skull. This and other traits nest it with the basal scleroglossan, Calanguban in the large reptile tree.

JZC Bu1803
(3.2cm snout vent length) Is more thoroughly covered in fine scales. Nevertheless, a large enough list of traits was gleaned from the photo to nest JZC Bu 1803 with the basal scleroglossan, Calanguban (Simoes, Caldwell and Kellner 2014, Early Cretaceous), which is also the same size.

According to Daza et al. “The ventral scales are large, quadrangular, and arranged in regular transverse and longitudinal rows as in most living teiids and lacertids. Remarkable features of this specimen are the extremely long digits and claws.

“A digital endocast of the leftmaxilla exposed the entire tooth row section, a dentition including about 14 functional teeth, and an estimated 19 tooth loci. Tooth attachment is pleurodont and morphology is heterodont, with an abrupt transition from conical and recurved teeth (first 12) to tricuspid (with divided crowns, last 7). Tricuspid (or triconodonot) dentition is widespread among squamates. Among extant groups, the combination of anterior fang-like and posterior tricuspid teeth with parallel margins, where mesial and distal cusps are shorter than the main apex,most closely resembles that of lacertids and teiids.”

Figure 5. Calanguban nests as a sister to JZC Bu 1803 in the large reptile tree.

So this find
extends the range of the Calanguban clade from SW Europe across the then narrow Atlantic to NE Brazil.

Figure 6. The two new and unidentified amber embedded specimens nest as squamates that already have names here.

The authors indicate
that more data will be forthcoming on these specimens. More can be seen on YouTube here.

References
Bolet A and Evans SE 2012. A tiny lizard (Lepidosauria, Squamata) from the lower Cretaceous of Spain. Palaeontology 55:491-500.
Daza J, Sanley EL, Wagner P, Bauer A and Grimaldi DA 2016. Mid-Cretaceous amber fossils illuminate the past diversity of tropical lizards. Science Advances 2(3): e1501080. DOI: 10.1126/sciadv.1501080
Simoes TR, Caldwell MW and Kellner AWA 2014. A new Early Cretaceous lizard species from Brazil, and the phylogenetic postion of the oldest known South American squamates. Journal of Systematic Palaeontology. http://dx.doi.org/10.1080/14772019.2014.947342

YouTube with rotating scans

 

 

 

 

The water-walker, the rib-glider and the frill-neck are all cousins!

Update March 5, 2016 with the addition of the cladogram from the large reptile tree.

Some reptile oddballs
>do< nest together. In this case, the extant “Jesus” lizard, Basiliscus, the rib-glider, Draco, and the frill-neck, Chlamydosaurus now nest together in the large reptile tree.

Figure 1. Draco volans. Note the anterior maxillary fangs, and the antorbital fenestra between the lacrimal and prefrontal, traits shared with Chlamydosaurus (Fig 2).

Draco (Fig. 1) and Chlamydosaurus (Fig. 2) are particularly interesting
as both share anterior maxillary fangs and an antorbital fenestra between the prefrontal and jugal (rather than between the lacrimal and maxilla as in other taxa with an antorbital fenestra). A long list of shared character traits unites these two still quite different lizards.

Figure 2. Chlamydosaurus also has anterior maxillary fangs and an antorbital fenestra between the prefrontal and lacrimal, as in Draco (Fig.1).

Basiliscus
is (at this point in the proceedings) a sister to the last common ancestor. But with that parietal crest, it has definitely evolved apart from the above two taxa for a long time.

Figure 3. Basiliscus, the “Jesus” lizard, does not share as many traits as Draco and Chlamydosaurus do, but is related, given the short list of Iguanids currently employed.

We’ve seen this before, 
where and when some odd little reptiles shared more traits with each other than with any other tested reptiles. Members of the Fenestrasauria (Cosesaurus, Kyrgyzsaurus, Sharovipteryx, Longisquama, and pterosaurs) also include bipeds with dorsal frills. One of them also glided with outstretched ribs and legs, although distinct from Draco. In Sharovipteryx, the hind legs were much longer than the ribs.

Figure addendum 1. Cladogram of the Iguania, the sister taxa of the Scleroglossa, both members of the clade Squamata, a subset of the clade Protosquamata, the sister taxon to the Tritosauria. Many more scleroglossans are shown in the large reptile tree at ReptileEvolution.com.

We don’t have 
close prehistoric relatives for Draco or Chlamydosaurus yet. So at this point the evolution of rib-gliding or frill-spreading is not yet a gradual demonstration. But the other shared traits are, to my knowledge, unique synapomorphies.

I will update the cladogram this weekend.

 

The other Slavoia and the holotype

Earlier we looked at the basal amphisbaenid, Slavoia darevskii (Fig. 3 below, Talanda 2015).

I just read about the holotype (Sulminski 1984) and at least 45 other specimens attributed to Slavoia, like this one (Fig. 1, ZPAL MgR-III/77, Campanian, Late Cretaceous). Six of the 46 skulls are associated with postcranial skeletons, like the holotype, Fig. 2, ZPAL MgR-I/8). If you think this skull looks like Macrocephalosaurus, you’re not the only one.

Slavoia specimen ZPAL MgR III/77, one of 46 skulls,  nests not with amphisbaenids, but with Macrocephalosaurus, a contemporary from the same horizon. Talanda reports, “The specimen has only half of the elements visible in this drawing. The skull roof and the middle part are not preserved.”

Sulminski nested Slavoia with scincomorphan lizards, but he reported, “It is interesting that described here lizard displays some characters similar to macrocephalosaurid and polyglyphanodontid  species discovered in the same localities of Mongolia. This concerns also particularly the structure of the temporal region, palatal construction and in number of teeth.”

The #77 and #8 specimens nested with macrocephalosaurs in the large reptile tree.

On the other hand,
the #112 specimen nested at the base of the amphisbaenids, as we learned earlier. So the #112 specimen needs a new generic name, or there are other issues that need be dealt with.

Dragging
the amphisbaenid #112 specimen over to the macrocephalosaur specimens adds 17 steps to the most parsimonious tree score. That’s a very low number considering that there are only 17 taxa separating the macrocephalosaurs from the amphisbaenids in the large reptile tree. So, there is a bit of convergence going on here between the macrocephalosaurids and amphisbaenids. The authors note all the skulls vary in size and shape, which they attribute to ontogeny and intraspecific variation. And, of course, none are perfectly preserved. Talanda reports, “The [#77] specimen has only a half of the elements visible in this drawing. The skull roof and the middle part are not preserved.”

Figure 2. The holotype of Slavoia (#8) compared to the lateral view skull (#77). While larger, the #77 skull is relatively shorter. These two nest together in the large reptile tree along with macrocephalosaurids. Note the large size of the limbs.

Does this represent a solution?
Sulimski (1984) recognized the similarity between his skinks and macrocephalosaurids. Talanda (2015) considered his specimen a basal amphisbaenid, a clade derived from skinks in the large reptile tree, but Talanda nested his amphisbaenids between Cryptolacerta and Dibamus + snakes. So there is disagreement here.

Figure 3. The #112 specimen from Talanda 2015 which both he and I nested as a basal amphisbaenid. Note the similarity to macrocephalosaurids (above). The teeth appear to be more robust here, as they are in the palate view specimens that have more of an amphisbaenid palate. I don’t see large limbs here, but limb size varies in the amphisbaenids.

Phylogeny is sometimes simple and straightforward.
Sometimes it is not.

This case shows the importance
of using specimen-based taxa in analyses, not specific, generic or suprageneric taxa. It would not be okay to take the best traits from several Slavoia specimens because some may not be Slavoia specimens! This case also highlights a need to determine where every one of these varied Slavoia specimens do nest. And it will be okay if some are lumped while others are split.

The limbs are large in the #8 specimen, but are not visible in the #112 specimen. In amphisbaenids limbs, even in basal taxa, can be vestiges, but not vestiges in the very derived Bipes.

We all have a lot to learn here. It’s not all set in stone.

References
Sulimski A 1984. A new Cretaceous scincomorph lizard from Mongolia. Palaeontologia Polonica, 46, 143–155.
Talanda M 2015. Cretaceous roots of the amphisbaenian lizards. Zoologica Scripta. doi:10.1111/zsc.12138

Reconstruction from jumbled scraps: the squamate, Kuroyuriella

Figure 1. The skull of Kuroyuriella (represented by two specimens of different size) reconstructed from bone scraps (above), most of which are layered on top of one another. Not all elements are identified, but enough are known to score and nest this taxon with Ophisaurus.

When provided disarticulated scraps,
start with the easy bones, then fill in the gaps in the puzzle. Sometimes, as in Kuroyuriella mikikoi (Evans and Matsumoto 2015, Early Cretaceous), there are enough parts to more or less recreate the skull most similar (among tested taxa in the large reptile tree) to that of Ophisaurus and basal to Myrmecodaptria and Cryptolacerta. Evans and Matsumoto nested Kuroyuriella  between Huehuecuetzpalli and the suprageneric clade Rhynchocephalia, both well outside the Squamata.

From the online paper:
“Together, SBEI 1510 and SBEI 1608, as type and referred specimen, characterize Kuroyuriella mikikoi as a small lizard having paired frontals with deep subolfactory processes; a median parietal without a parietal foramen, with sculpture of low relief, and with lateral shelves that restricted the adductor muscle origins to the ventral surface; upper temporal fenestrae that were at least partially closed by expanded postorbitofrontals; an unsculptured maxilla with a strongly concave narial margin; a large flared prefrontal; and a slender, relatively small pterygoid. In the shallow lower jaw, the teeth are closely packed, cylindrical, and pleurodont with lingual replacement; a subdental ridge is present; the dentary bears a tapering coronoid process that braces the coronoid, and has a posterior extension with a curved free margin; the surangular, angular, and splenial are all present and the surangular is shallow; the adductor fossa is open but not expanded; and the articular surface is asymmetrical.

In order to explore the affinities of Kuroyuriella mikikoi, it was coded into the matrix of Gauthier et al. (2012), as extended by Longrich et al. (2012) (184 characters coded out of 622, 70.4% missing data),

The consistent placement of Kuroyuriella on the squamate stem is problematic and probably artifactual, but whether the weighted analysis is giving a more accurate placement is uncertain. Of the derived character states possessed by Kuroyuriella, 76 [1] (postorbital partly occludes upper temporal fenestra), 364 [1] (dentary coronoid process extends beyond level of coronoid apex), 367 [2] (coronoid process of dentary overlaps most of anterolateral surface of coronoid), and 369 [2] (dentary terminates well posterior to coronoid apex) provide some support for placement of Kuroyuriella on the stem of scincids, and 129 [1] (prefrontal extends to mid-orbit), 104 [1] (absence of parietal foramen) and 385 [1] (posterior mylohyoid foramen posterior to coronoid apex) would be consistent with that placement. However, given the considerable difference between the results using equal weighting and Implied Weighting, Kuroyuriella remains incertae sedis pending recovery of more complete material.”

Figure 2. Ophisaurus, the extant glass snake or legless lizard is close to Kuroyuriella in the large reptile tree.

Here
Ophisaurus (Fig. 2) and Kuroyuriella both nest will within the Squamata, not ouside. of it in the large reptile tree. Reconstruction of the skull helps to ‘see’ this lizard as it was. I can’t imagine how difficult it would be to do try to establish traits  with a jumble of disarticulated bones.

As you’ll see, I think the parietal foramen was present. The parietal may have had longer posterior processes, now broken off.

References
Evans SE and Matsumoto R 2015. An assemblage of lizards from the Early Cretaceous of Japan. Palaeontologia Electronica 18.2.36A: 1-36
palaeo-electronica.org/content/2015/1271-japanese-fossil-lizards
http://palaeo-electronica.org/content/2015/1271-japanese-fossil-lizards

Urumqia – a very basal lepidosauromorph

Figure 1. Urumqia liudaowanensis (Zhang et al. 1984) ~20 cm snout-vent length, Late Permian.

Here’s a gephyrostegid/basal amniote/basal lepidosauromorph
you may not have heard of. (Remember lepidosauromorphs in the large reptile tree constitute about half of all amniotes). It is considered China’s oldest known tetrapod.

Urumqia liudaowanensis (Zhang et al. 1984, Fig. 1) ~20 cm snout-vent length, Late Permian Lucaogou Formation), was originally considered a discosaurid seymouriamorph. Here it nests at the base of the lepidosauromorph reptiles. Shifting Urumqia to the discosaurid seymouriamorphs adds 39 steps to the large reptile tree.

Derived from 
Gephyostegus bohemicus, Urumqia was basal to Bruketererpeton, Thuringothyris, and all lepdiosaurs, turtles, diadectids, pterosaurs and other various lepidosauromorphs starting with Saurorictus and Cephalerpeton. Phylogenetically Urumqia must have made a first appearance in the Viséan (335 mya, Mississipian, Carboniferous) despite its late appearance in the Late Permian (255 mya).

Figure 2. Basal amniotes to scale. Click to enlarge. Urumqia nests on the right hand column with Cephalerpeton and Thuringothyris.

Distinct from G. bohemicus,
Urumqia had shorter limbs, longer (but not long) posterior dorsal ribs and a robust tail with elongate caudals. The palate included a suborbital fenestra. The cheek may have included a small lateral temporal fenestra convergent with others. The carpals and tarsals were poorly ossified.

Figure 3. Extant lizards, A. gravid, B. in the process of laying eggs, C. with egg clutch.

Notably
the posterior dorsal ribs were much shorter than the gastralia. So the gastralia create a wide posterior torso, ideal for carrying large amniote eggs (Fig. 3), as we learned earlier.

The new topology of basal reptiles
is based on the inclusion of several more species based taxa not previously considered. This new topology show that synapsids were not the first clade to branch off. Rather all taxa closer to archosaurs (here considered the new Archosauromorpha) split from all taxa closer to lepidosaurs (here considered the new Lepidosauromorpha) at the onset of the Reptilia (=Amniota).

References
Zhang F, Li Y, and Wan X. 1984. A new occurrence of Permian seymouriamorphs in Xinjiang, China. Vertebrate Palasiatica22(4):294-306.