332 A SAUROPODOTJS DINOSAUR FROM MADAGASCAR. [Aug. 1 89 5. 



which probably belongs to the anterior part of the series, and is broken 

 off along the line of the floor of the neural canal. The lateral cavity, 

 although incomplete, appears to be oval in form, with its longer 

 axis longitudinal ; and its lower outer margin is not produced 

 upwards to form a pocket. The centrum is shorter and wider than 

 in the cervicals, with its lower surface broad and flat ; in all of 

 which respects it agrees exactly with the corresponding vertebrae of 

 Jlojolosaurus. The anterior terminal convexity is also less strongly 

 marked than in the cervicals. In length this specimen measures 

 7| inches, while its distal width is about 7| inches. 



The centrum and base of the neural arch of the very large 

 vertebra shown in fig. 4, from the presence of a small lateral im- 

 mediately beneath the base of the transverse process, I take to be 

 the last lumbar rather than the first caudal. The centrum is very 

 short and wide, with its anterior terminal face somewhat saddle- 

 shaped from above downwards, and the posterior extremity flat. 

 The length of the centrum is 5 inches, and its posterior width 

 upwards of 10 inches. • 



The sacrum is represented by two fragments, showing the anterior 

 and posterior terminal surfaces, and probably belonging to a single 

 individual. From these specimens it is probable that there were at 

 least four segments in this portion of the vertebral column. The 

 terminal faces are quite flat, and there are no lateral pits to the 

 centra. 



The centrum of a first caudal vertebra of large size is charac- 

 terized by its extreme shortness and width : the anterior face being 

 saddle-shaped from above downwards, and the posterior cupped. 

 The length is 4|, and the width 10 inches. 



Both in form and size this vertebra closely resembles the anterior 

 caudals of Pelorosaurus Leedsi, from the Oxford Clay of Peter- 

 borough, although the latter are abnormally shortened from the 

 effects of pressure. In this, as well as in the more posterior caudals, 

 there are no lateral pits. 



From among several examples of vertebras occupying a more 

 posterior position in the caudal series, the one represented in fig. 5 

 (p. 334) is selected for illustration. Relatively longer than the last, 

 this specimen has its anterior face flat and heart-shaped, while the 

 hinder face is somewhat cupped. Chevron-facets are shown on 

 both the front and hind borders of the inferior surface, indicating 

 that the chevrons articulated with two adjacent vertebrae. The 

 posterior caudals, of which there are several more or less imperfect 

 examples, have the general form common to sauropodous dinosaurs, 

 with flat terminal faces. 



Since the remains of the pectoral and pelvic girdles, as well as 

 the limb-bones, are very imperfect and broken, it seems unnecessary 

 to describe them, more especially as bones like the femur do not 

 generally afford any very well-marked generic characters. It may 

 be mentioned, however, that the two extremities of a femur indicate 

 a bone apparently closely resembling in size and form the femur of 

 Cetiosaurus oxoniensis. 



