MATERIALS AND METHODS

The heads of two alcohol-preserved specimens of S. crocodilurus were scanned at The University of Texas at Austin High-Resolution X-ray Computed Tomography (CT) Facility (UTCT). The first specimen (Field Museum of Natural History, FMNH 215541; 136.5 mm snout-vent length [SVL]) is an adult collected from Jinxin County, Guangxi Province, China, in June 1978. The original data set of the adult consists of 465 CT slices taken along the coronal or transverse axis with both a slice thickness and interslice spacing of 78.4 µm. The field of reconstruction is 30.0 mm with an image resolution of 1024 x 1024 pixels, resulting in an interpixel spacing of 29.3 µm. The second specimen (Texas Natural History Collection, Texas Memorial Museum, TNHC 62987; 59.1 mm SVL) is a juvenile specimen for which no locality information in known. The original data set of the juvenile consists of a total of 516 coronal CT slices, with both the slice thickness and interslice spacing equaling 29.0 µm. The field of reconstruction is 12.5 mm with an image resolution of 512 x 512 pixels, resulting in an interpixel spacing of 24.4 µm. The difference in resolution between the juvenile and adult datasets is due to technological upgrades of the UTCT scanner between April 2000, when the juvenile was scanned, and March 2004, when the adult was scanned. That difference in resolution, as well as the lesser degree of ossification, results in poorer quality of the juvenile dataset.

The original coronal CT slices for both juvenile and adult were resliced along the horizontal and sagittal planes and then used to create three-dimensional (3-D) digital reconstructions of the skulls. The skulls were then segmented into constituent parts (e.g., dermatocranium, splanchnocranium, neurocranium) using VG Studio Max© (Volume Graphics). We performed additional segmentation on the braincase of both the juvenile and the adult so that individual bones could be studied and figured in isolation. We also rendered a digital endocast of the cavities of the inner ear of the adult to study its structure.

Certain areas (e.g., portions of the prootic covering the anterior semicircular canal, portions of the parasphenoid rostrum) appear to be absent in the three-dimensional reconstructions. This is because the grayscale (= density) values of these extremely thin regions are similar enough to the surrounding soft tissues that they ‘disappear’ when the soft tissues are digitally rendered transparent.

The text and figures from this paper are complemented by the following 3-D reconstructions animated to rotate around the x and y axes: adult (FMNH 215541) and juvenile (TNHC 62987) braincases (Appendix 1 and Appendix 2), skull and associated osteoderms of the adult (FMNH 215541) (Appendix 3), and the digital endocast of the left inner ear cavities of the adult (FMNH 215541) (Appendix 4). Slice-by-slice animations along the coronal axis also are provided for both the adult (FMNH 215541) and juvenile specimens (TNHC 62987) (Appendix 5 and Appendix 6).

Additional digital animations of skulls and skull elements for both the adult and juvenile specimens were scanned and posted at DigiMorph Shinisaurus crocodilurus and DigiMorph Shinisaurus crocodilurus (juvenile). These images include 3D reconstructions (animated to rotate around the three orthogonal axes) of the entire skull, articulated braincase, isolated orbitosphenoids, isolated sphenoid, isolated basioccipital, isolated supraoccipital, isolated prootic, isolated otooccipital, and isolated stapes. Additional imagery for the juvenile includes slice-by-slice animations along the three orthogonal axes, 3-D cutaway animations along the three orthogonal axes, and an inspeCTor Java viewer applet. Additional animations for the adult specimen include 3-D digital endocast of the left inner ear cavities rotating around the three orthogonal axes, and 3-D reconstructions of the skull with osteoderms digitally removed rotating around the three orthogonal axes.

In addition to the CT data sets, we referred to two dry skeletal preparations of S. crocodilurus. A specimen from the University of California Museum of Vertebrate Zoology (MVZ 204291) includes an articulated skull of an adult of unspecified SVL. A juvenile specimen from the Quaternary Sciences collection at Northern Arizona University (NAUQSP 17563) has a disarticulated skull and SVL of 59 mm.

Unless otherwise specified, terminology follows that used by S.E. Evans in a forthcoming paper to be published in Volume 20 of Biology of the Reptilia. Two exceptions are: 1) we do not use directional terms in the naming of processes and 2) we adopt a standard usage for the terms ‘foramen’ and ‘fenestra’. We use ‘foramen’ for a penetration within a bone and ‘fenestra’ for openings between bones; a notable exception to this is our retained usage of the commonly accepted ‘foramen magnum.’