Annals of the Transvaal Museum 8i 
very probable that the third sacral rib was too short to reach the ileum, not- 
withstanding normal backward convergence of the ilea. However, this may 
be explained in the same way as certain differences in Theropod sacra by 
v. Huene ( 3 1 , p. 281). As appears from many loose sutural connections, our 
animal is a young individual, and has not yet reached the stage of life in which 
it possesses a true third sacral vertebra. 
Caudal Vertebrae. 
Of the caudal vertebrae only twelve have been preserved. The first two 
caudals are present. The first caudal lies immediately behind the last sacral 
and the second caudal behind the first. Then follows a large gap. After this 
comes a series of eight vertebrae. Two more vertebrae can be recognised in 
a jumbled up mass at the end of this series. 
The centrum of the first caudal has a length of 15 mm. That of the second 
cannot be accurately measured, because it has split obliquely. It is also diffi- 
cult to measure the height of the centra, but the dimension seems to be slightly 
less or equal to the length. The lower portion of the hinder articulation surface 
of both centra has been obliquely truncated to form an articulation surface 
for the haemapophysis. In the second caudal it is clearly visible that the 
lower portion of the -front articulation surface bends slightly backwards. The 
transverse processes of the right side have both been disjointed and their 
impressions lie at a distance of about 2-5 cm. from their respective vertebrae. 
They lie in a row with the second and third sacral ribs. The process of the 
first caudal has a length of 18 mm. and a breadth of 10 mm. These measure- 
ments of that of the second caudal are respectively 16 mm. and 10 mm. Both 
processes show great resemblance to the third sacral rib. 
The series of eight vertebrae is from the middle region of the tail. The 
centra are very long and low and they all show their lower surface and that 
of their right transverse process. The dorsal spines of all these vertebrae are 
embedded in the rock and invisible. The centra are so much flattened, that 
it is impossible to give any detail. The articulation surfaces of the centra 
project beyond their sides and lower surface. The transverse processes are 
remarkable for their broad attachment to the centrum. In the first vertebrae 
of the series they are practically attached to the whole length of the centrum, 
while in the last the attachment is still about two-thirds of this length. The 
distal three-fifths of the process has parallel sides, but those of the proximal 
end diverge towards the centrum. The process was apparently supported by 
lateral ridges. There are indications that the front lateral ridge of the process 
was in touch, in all the vertebrae of the series, with the praezygapophysis. 
Haemapophyses are attached to all these vertebrae and still in situ. 
Measurements from the first vertebra in the series to the last : 
Length of centrum 17, 17, 18, 18, 18, 18, 19, 19 mm. 
Length of transverse process 18, 19, 18, 15, 15, 15, — , 12 mm. 
Length of haemapophyses 40, 37, 31, 31 mm. 
Breadth of their distal ends 7, 7, 7, 7 mm. 
In front of the first vertebra of the series lies another haemapophysis, which 
through its position gives the impression of having belonged to the first pre- 
ceding vertebra. This haemapophysis has a length of 42 mm. Four or five 
other haemapophyses are scattered about, one of which is perfect. This has 
a length of 45 mm. The distal end of all these haemapophyses is broad and 
very thin. The proximal end, distal to the foramen, is narrower and thicker. 
The foramen of the longest haemapophysis has a height of 7 mm. and a breadth 
1 These numbers refer to the list of literature at the end of the article. 
