FIGURE 1. Tail skeletons of a theropod dinosaur and a pterosaur, showing the bony processes that restrict lateral bending and therefore prevent tail coiling; anterior is to the left. 1.1. The theropod dinosaur Allosaurus fragilis. 1.2. The long-tailed pterosaur Rhamphorhynchus gemmingi, modified from a published illustration Wellnhofer (1975, Figure 7). 1.3. A single tail vertebra and hemal arch of R. gemmingi. prz = prezygapophysis, ha = hemal arch.
FIGURE 2. The draco of Pliny: African and Indian pythons. 2.1. African rock python (Python sebae). 2.2. Indian rock python (Python molurus).
FIGURE 3. Snakes in Roman wall paintings, with the crests and wattles of roosters, from two houses in Pompeii (from Grant, 1979, unnumbered figures).
FIGURE 4. Examples of the drak?n in ancient Greek art depicting scenes from mythology. Note that they are all snakes, and that the ancient Greeks often added beards and occasionally rooster crests to snake depictions. 4.1. The drak?n of the Hesperides, which was slain by Herakles, from a Greek vase from ca. 500 B.C. (Carpenter, 1991, Figure 212). 4.2. The drak?n of the Hesperides, from a Greek vase from the fifth century B.C. (Carpenter, 1991, Figure 213) 4.3. The drakontes that made up the belt of a gorgon, from a Greek jug from the sixth century B.C. (Woodford, 2003, Figure 95). 4.4. The drakontes that made up the belt of a gorgon, from a Greek temple relief from the sixth century B.C. (Carpenter, 1991, Figure 155). 4.5. The drak?n that was the tail of the Chimaera, from a relief on a shield band panel from the sixth century B.C. (Carpenter, 1991, Figure 164). 4.6. The drak?n that was the tail of the Chimaera, from a Greek relief from the fifth century B.C. (Buxton, 2004, unnumbered figure). 4.7. The drak?n slain by Kadmos, from a Greek vase from the fifth century B.C. (Buxton, unnumbered figure). 4.8. The drakaina slain by Apollo, from a Greek coin from the fifth century B.C. (Carpenter,1991 Figure 104).
FIGURE 5. The krokodilopardalis in the Nile mosaic of Palestrina, compared to an otter and to three representative theropod dinosaurs. Note the resemblance between the krokodilopardalis and the otter, and the marked difference between the body plan of the krokodilopardalis and the theropod body plan. 5.1. The krokodilopardalis (Meyboom, 1995, Figure 18). 5.2. An otter (Lontra canadensis). 5.3. Reconstruction of he dromaeosaurid theropod Microraptor gui. 5.4. Skeleton of the carnosaurian theropod Acrocanthosaurus atokensis. 5.5. Reconstruction of the compsognathid theropod Compsognathus longipes.
FIGURE 6. An alleged dinosaur (actually a crocodile) in a Pompeii wall painting, compared with other crocodiles in Roman wall paintings. Note that in all cases, the crocodiles are not portrayed with great realism but are dumpy and almost cartoonish. 6.1. The alleged dinosaur (Zillmer, 1998, Figure 91). 6.2. Crocodile in a Roman relief from the first century B.C (Meyboom, 1995, Figure 38). 6.3. Crocodile in a Roman mosaic from the first century B.C (Meyboom, 1995, Figure 28).
FIGURE 7. A long-tailed pterosaur, a short-tailed toothed pterosaur, and a short-tailed toothless pterosaur. These pictures show that pterosaurs of all three sorts do not have the form of a water snake, and that in all three the body plan is more birdlike than snakelike, even without wings. 7.1. Scaphognathus crassirostris, skeleton as preserved and illustrated by Goldfuss (1831). 7.2. Haopterus gracilis, reconstructed according to a complete skeleton. 7.3. Sinopterus dongi, reconstructed according to a complete skeleton. 7.4. The previous three pterosaurs, with wings removed, to show that they are not snakelike at all and therefore Herodotus' description of flying serpents as "having the form of the water snake" is inapplicable to pterosaurs; the bodily form of S. crassirostris is reconstructed according to the complete skeleton of a juvenile (Wellnhofer, 1991, unnumbered figure), with the proportions altered to match those of the adult of a closely related species: Rhamphorhynchus gemmingi.
FIGURE 8. Two alleged representations of the pterosaur Scaphognathus crassirostris, and scenes of snakes pulling chariots in classical art. Note that the chariot-pulling "pterosaurs" are winged snakes with crests and wattles that resemble those of the chariot-pulling snakes in 8.5. Note also the lack of resemblance between S. crassirostris itself (in Figure 7.1) and the alleged S. crassirostris representations in 8.1 and 8.2. 8.1. A pair of alleged S. crassirostris pulling the chariot of Triptolemos in a Roman-Alexandrian coin (Goertzen, 1998, Figure 7). 8.2.Head of an Etruscan statue, allegedly of S. crassirostris (Goertzen, 1998, Figure 11). 8.3. Winged snakes pulling Triptolemos' chariot, on a Thracian coin from the second or third century A.D. (Sayles, 1998, unnumbered figure). 8.4. Wingless snakes pulling Triptolemos' chariot, on a Roman sarcophagus from the third century (Robert, 1919, Figure 433). 8.5. Wingless snakes pulling the chariot of the Roman goddess Ceres, on a Roman coin from the first century B.C. (Spaeth, 1996, figure 5).
FIGURE 9. An alleged pair of Tanystropheus (actually a mythical sea monster called a cetus) in a second-century Roman mosaic in Lydney Park, England, compared with Tanystropheus and other examples of cete. Note that different artists portrayed the cetus with different, creative flourishes of anatomical interpretation, but the basic form of the cetus remains relatively uniform: a long-necked, long-eared, fluke-tailed marine creature with two fore-paws or fore-flippers and no hind appendages, often with a coiled tail and a tufted nose. 9.1. The Lydney Park creatures (Taylor, 1987, unnumbered figure). 9.2. the Triassic reptile Tanystropheus longobardicus. 9.3. Cetus in a fourth-century mosaic from Syria (Dunbabin, 1999, Figure 173). 9.4. Cetus in a fourth-century Roman mosaic (Poeschke, 2010, Figure 1). 9.5. Cetus on a third-century Roman sarcophagus (Jensen, 2004, Figure 29). 9.6. Cetus on a fourth- or fifth-century tray from Carthage (Lazaridou, 2011, Figure 14). 9.7. Cetus in a first-century Roman fresco (Woodford, 2004, figure 8.9).
