Figs. 2B, 4, 5B

= Tarbosaurus efremovi Maleev 1955b

= Gorgosaurus lancinator Maleev 1955b

= Tarbosaurus bataar (Maleev) 1955 b

Holotype: PIN 551-1 partial skull from the Nemegt Formation, Mongolian People’s Republic.

Diagnosis: Angular terminates anterior to the surangular fenestra;surangular fenestra proportionally smaller than in Tyrannosaurus rex; surangular fenestra proportionally smaller than in T. rex.

Stratigraphic and Paleobiogeographic Distribution: Nemegt Formation and Upper White Beds of Khermeen Tsav, Mongolian People’s Republic; possibly the Subashi Formation,Xinjiang,People’s Republic of China.

Discussion: Maleev (1955a,b; 1974) recognized three genera and four species of tyrannosaurids from Mongolia: Tyrannosaurus bataar, Tarbosaurus efremovi, Gorgosaurus lancinator and Gorgosaurus novojilovi. Rozhdestvensky (1965),however,argued that most of the differences cited by Maleev were ontogenetic, and that only a single species was represented. Rozhdestvensky also argued that this single species was intermediate between Albertosaurus and Tyrannosaurus and, thus, represented a distinct genus for which the name Tarbosaurus was available. Accordingly, Rozhdestvensky referred all Mongolian tyrannosaurid to Tarbosaurus bataar.

Re-examination of Maleev’s figures (Maleev, 1955a, 1955 b; 1974),as well as photographs of these specimens and of those collected by the Polish palaeontological expeditions to Mongolia, reveal that Rozhdestvensky erred in a number of points in his analysis. The quadratojugal process of the jugal in T. bataar reaches the posterior rim of the lateral temporal fenestra as it does in T. rex. The tetragonal-shaped ascending process of the jugal in T. rex,this observation is based on AMNH 5027 (Osborn, 1912). A new skull of T. rex, LACM 23844,has a tapering ascending process as in PIN 551-3, thus showing the shape of this bone is variable. Development of the ascending process into the orbit is variable as indicated by the holotype T. bataar, PIN 551-1, which shows a slight development of the jugal within the orbint. This makes it intermediate in shape between AMNH 5027 and PIN 551-3.

Another taxonomical difference Rozhdestvensky stated was the smaller size of the Mongolian tyrannosaurid. However,the holotype skull, PIN 551-1,has premaxillary to occipital condyle length of 1220 mm (Maleev, 1955a, 1974). This compares to 1210 mm for AMNH 5027, T. rex (Osborn, 1912).

Thus,contrary to Rozhdestvensky’s(1965) claims, T. bataar is not intermediate between Albertosaurus and Tyrannosaurus, but is generically indistinguishable from Tyrannosaurus. It is remarkable that of the five skulls of T. rex known all are adults (however see below). This is not the case with T. bataar. Figure 4 shows a growth series of skulls from the smallest with a premaxillary-occipital condyle length of 825 mm to the largest with a premaxillary-occipital condyle length of 1220mm. The general ontogenetic trend is a deepening of the skull, shortening of the muzzle,and an increase in the rugosity of the region dorsal to the orbit. As may be seen, the smaller skulls resemble that of Albertosaurus (compare Figs. 4 C, D with Fig. 2 E),especially because of the longer muzzle. Thus,it is understandable that Maleev referred these skulls to Gorgosaurus (= Albertosaurus). However,the well-developed stereoscopic vision (due to the differentiation of the muzzle from the cranium) separates these skulls from Albertosaurus. Variation in the skulls is apparent in the posterior portion of the lateral temporal fenestra where the squamosal may butt against the quadratojugal (Fig. 4 B and D) or there may be an overlap (Fig. 4 C). A similar pattern is seen in T. rex as well (Carpenter,in press).

There has been some confusion about the astragulus in Tyrannosaurus. Welles and Long (1974) figured an element associated with AMNH 5827 that they ascribed as a Tyrannosaurus astragulus. This element, however, is unlike any theropod astragulus known,and is certainly very different from the astragulus associated with MOR 009, Tyrannosaurus rex. This astragulus resembles that of Albertosaurus as would be expected. Comparison of the original drawings by Christman of the element figured by Welles and Long show that it is,in fact,a medial and dorsal view of the right quadrate of AMNH 5027 (Fig. 3).

At this time the specific differences between the postcrania of T. rew and T, bataar have not been established. This is due to the considerable amount of variation among the T. rex specimens (Carpenter,in press). Some of the differences may be individual, but most appear to be sexual. Until the range of this variation can be determined, it is not certain what postcranial characters separate T. rex from T. bataar.

With regard to the maxilla from the Tornillo Formation of Texas ascribed by Lawson (1976) to T. rex,it does not belong to Tyrannosaurus (Carpenter,in press), but another short-faced theropod.

A description of the postcrania of Tyrannosaurus is in preparation.