Move from named bones and CT volumes to competing matrices, clock calibrations and explicit Mosasauroidea–Mosasauridae boundaries without turning separate fossils into an ancestor ladder.
Each step names its evidence claims, with references below. Interactive Explorer states are available in the full application; the narrative remains an editorial synthesis.
NHM PV R 4949 shows successive bone deposition in a broadened Eunotosaurus rib, while other specimens preserve shoulder anatomy. The tissue is observed; turtle-stem placement, homology and function are tested interpretations.
Separate preserved tissue from a character-homology model and from direct ancestry.
claim:event:eunotosaurus-rib-histology-shell-modelThe Vellberg Pappochelys sample combines broad ribs and robust paired gastralia but lacks a fused plastron and complete carapace. Serial fusion is a homology hypothesis, not a frame in a literal ancestor sequence.
A multi-specimen reconstruction is not one individual transforming into a later taxon.
claim:event:pappochelys-gastralia-shell-mosaicThree Odontochelys specimens preserve a developed plastron with expanded ribs and neural plates but no complete carapace. A plastron-first sequence and marginal-marine life remain interpretations, not anatomy seen changing through time.
Shell elements, environmental setting and a proposed evolutionary order are distinct claim layers.
claim:event:odontochelys-plastron-dorsal-shell-mosaicCaribemys holotype MNHNCu P-3209 anchors a 155.6 Ma hard minimum for crown Testudines in one best-practice analysis. Its much older posterior estimate comes from a relaxed clock, topology and priors.
A fossil minimum, a maximum prior and a posterior node age are three different quantities.
claim:event:caribemys-crown-turtle-calibrationMicro-CT reveals the cranial mosaic of Taytalura holotype PVSJ 698. Multiple analyses place it on the lepidosauromorph stem, but a single skull does not supply whole-body anatomy, direct ancestry or a global range.
The scan volume is specimen evidence; the tree position is an analytical output.
claim:event:taytalura-stem-lepidosaur-skullCT anatomy of Megachirella PZO 628 was combined with fossil morphology and living molecular partitions. The analysis recovers a stem squamate and estimates deep divergences, but neither output is a directly observed crown FAD.
Fossil anatomy, molecular constraint and clock time remain separately labelled.
claim:event:megachirella-ct-stem-squamateSynchrotron scans of Bellairsia NMS G.2022.1.1 reveal crown-like and retained states in one Jurassic skeleton. Stem placement and proposed cranial motion depend on matrices and functional interpretation.
Completeness improves observation but does not remove topology or function uncertainty.
claim:event:bellairsia-synchrotron-stem-squamateTeams disagree over bones, referrals and character scores for Cryptovaranoides NHMUK PV R36822, producing crown-squamate and non-squamate trees. The atlas therefore keeps the topology contested instead of inventing a secure Triassic crown FAD.
COL26.8 assigns 12,622 accepted living species to this package as a nomenclatural routing snapshot; it neither resolves fossil topology nor measures dossier, translation, media or review maturity.
claim:event:cryptovaranoides-competing-topologiesExplicit definitions make Mosasauroidea broader than node-defined Mosasauridae. Dallasaurus retains plesiopedal limbs, so family placement and full paddle specialization remain separate questions.
The Mosasauroidea navigation edge is non-taxonomic for discovery; basal topology remains analysis-sensitive.
claim:event:mosasauroid-clade-boundaries