Fig. 7. Vertebrae of Eusthenopteron (×1) and Ichthyostega (×2/3, after Jarvik, 1952), Trimerorhachis (×1-1/2, after Case), and Amphibamus (×10, after Watson, 1940) in lateral and end views; the two lower right-hand figures are from Watson (1940, as Miobatrachus); the lower left is from a cast of the "Miobatrachus" specimen in Chicago Natural History Museum, No. 2000, in the presacral region (original, ×10).
Fig. 7, lower right, is Watson's (1940) illustration of the anterior trunk vertebrae of Amphibamus (Miobatrachus), in which the intercentrum is shown as a single median piece. Fig. 7, lower left, shows two of the more posterior trunk vertebrae seen as impressions in a cast of the type of "Miobatrachus romeri;" evidently the inter-centra were paired at about the level of the 16th vertebra, and relatively large. Gregory's (1950) figure of the type specimen of "Mazonerpeton" (also equivalent to Amphibamus) shows the anterior trunk vertebrae in relation to the ribs essentially as they appear to me in the cast of Miobatrachus, and rather differently from Watson's figure of the latter. Gregory is probably right in considering the specimens to represent various degrees of immaturity. So far as present information goes, then, the vertebrae of salamanders and frogs show no clear evidence of derivation from those of any particular group among the early Amphibia, but their features are not inconsistent with a simplification of the pattern of Temnospondyli.
Fig. 8. Pectoral girdles of Protobatrachus (after Piveteau, 1937), Notobatrachus (after Stipanicic and Reig, 1956), Ascaphus (after Ritland, 1955 a) and Rana (original); all ×2. For explanation of abbreviations see Fig. 3.
PECTORAL GIRDLE
Hecht and Ruibal (Copeia, 1928:242) make a strong point of the nature of the pectoral girdle in Notobatrachus, as described recently by Stipanicic and Reig (1955, 1956) from the Jurassic of Patagonia, and quite rightly recommend that the significance of the arciferal and firmisternal types of girdle be restudied. That of Notobatrachus is said to be firmisternal; in view of the arciferal condition in the supposedly primitive Leiopelma, Ascaphus, Bombinator, etc., this comes as a surprise. Is the firmisternal girdle, as seen in Rana, Bufo, and others, actually the ancestral type, and has the arciferal been derived from something like this?
In the figures given by Stipanicic and Reig the ossified parts of the girdle are figured in detail (Fig. 8) and Reig's discussion of it is thorough. The decision to call it firmisternal was taken with some hesitancy, for no median elements are indicated, and the position and shape of those seen is closely similar to the ossified parts in Ascaphus and Leiopelma; there is no bony sternum or omosternum. It is safe to suppose that some cartilage lay in the midline between the clavicles and coracoids, but there is no evidence as to its extent, rigidity, or degree of overlapping if any. Apparently, then, there is not sufficient reason to infer that this Jurassic frog had a pectoral girdle comparable with the modern firmisternal type.
Piveteau (1955:261) remarks that the only living Anuran that can be compared usefully with Protobatrachus (Triassic) with regard to its pectoral girdle is Ascaphus. Again, the extent of cartilage in Protobatrachus (Fig. 8) can only be inferred, and there are no median elements. The agreement with Ascaphus includes the presence, in both, of a separate coracoid ossification situated posterior to the ossified "scapulocoracoid" (actually scapula). This ossification is evidently that shown in Notobatrachus as "coracoid." Direct comparison of the three genera with one another suggests that if we use the term arciferal for any, we should use it for all.