92 Annals of the Transvaal Museum 
The pubis of our form differs considerably from that of Thecodontosaurus 
antiquus ( 3 , PI. LXXXV), as far as the latter is known. In our form the 
lateral edge of the pubis is regularly concave, whereas in Thecodontosaurus 
antiquus its upper end is sigmoidal. There is also great difference in the shape 
of the pubic foramina. The shape of the proximal end of the ischium of 
Thecodontosaurus antiquus , as far as preserved ( 3 , p. 209, fig. 228), is quite 
different from that of our form, a difference which is best understood from a 
comparison of the figures. 
Another difference becomes conspicuous when the length of the humerus 
is expressed in lengths of dorsal vertebrae. Taking one of the hinder vertebrae 
v. Huene came to the following results ( 3 , p. 309) : In Thecodontosaurus anti- 
quus the humerus is about five times as long as the vertebra, and in The- 
codontosaurus skirtopodus about four and a half times. In our form the length 
of the eleventh dorsal vertebra is 29 mm. The length of the right humerus is 
93 mm., which means that the humerus is only 3*2 times as long as the vertebra. 
Therefore, the humerus of our form is relatively much shorter than that of 
Thecodontosaurus antiquus and of Thecodontosaurus skirtopodus. 
There is great difference between the ischium of Thecodontosaurus minor 
( 5 , p. 469) and that of our form. 
The points of difference enumerated above show sufficiently that our form 
does not belong to any of the known genera of the Anchisauridae. It, 
therefore, represents a new genus, for which I propose the name Aristosaurus 
n.g. with the species Aristosaurus erectus n. sp. 
Aristosaurus erectus is much more highly specialised than Thecodontosaurus, 
Ammosaurus, Anchisaurus and even than M assospondylus . The build of the 
pelvis, and especially the position of the ischium, shows adaptation to a 
usually bipedal mode of locomotion. The same may be concluded from the 
far forward position of the acetabular concavity in the ileum. The position 
of the trochanter quartus seems to be very low down on the femur. Its upper 
end is 41 mm. from the proximal end of the bone. The length of the trochanter 
is at least 18 mm. Therefore the lower end of the trochanter is situated at 
more than 59 mm. from the proximal end of the bone, which means very near 
to the middle of the femur. However, conclusions may not be drawn from 
this fact, because exact measurements cannot be obtained. 
The humerus is much shorter in relation to the body than in the other 
Anchisauridae. The anterior extremity is also relatively much shorter in 
relation to the posterior one than in all other Anchisauridae excepting 
Anchisaurus solus. As in the Plateosauridae the tibia of Aristosaurus is much 
longer than the humerus. This is also the case in Anchisaurus solus. In the 
other Anchisauridae it is the reverse. All this tends to show that Aristo- 
saurus is an Anchisaurid, highly specialised in the direction of the Plateo- 
sauridae, and of the bipedal mode of locomotion. 
Literature. 
1. Broom, R. On the South African Dinosaur ( Hortalotarsus ). Trans. S. Afr. Phil. 
Soc. vol. xvi. pp. 201-206, PI. hi. Cape Town, 1905. 
2 . Huene, F. von. Ueber die Dinosaurier der aussereuropaeischen Trias. Geol. u. 
Pal. Abh., herausg. v. E. Koken, N.F. Bd. vm. H. 2. Jena, 1906. 
3 . Huene, F. von. Die Dinosaurier der europaeischen Triasformation. Geol. u. 
Pal. Abh., herausg. v. E. Koken, Suppl.-Bd. 1. Jena, 1907-1908. 
4 . Broom, R. On the Dinosaurs of the Stormberg, South Africa. Ann. S. Afr. Mus. 
vol. vii. Pt. iv. No. 13, pp. 291-308, Pis. xiv-xvii. London, July 5th, 1911. 
5 . Haughton, S. H. A new Dinosaur from the Stormberg Beds of South Africa. 
Ann. and Mag. Nat. Hist. Ser. ix. vol. 2, No. 12, pp. 468-469. London, Dec. 1918. 
