Dr. Francis Baron Nopcsa — British Dinosaurs. 247 



iliac bone in PolacantJius, but I rather think we owe this simply 

 to the Professor's keen eye and intimate knowledge of Dinosaurian 

 structure. Until now, however, though already mentioned and 

 figured by Hulke, no attention has been paid to the fact that across 

 all the dorsal and sacral ribs ossified tendons (ligaments ossified) 

 extend which show a considerable amount of disturbance, and 

 thus clearly prove that they have not been co-ossified either with 

 the superimposed dorsal armour or with the underlying ribs. This 

 alone shows, apart from all other physiological considerations, that 

 right and left from the median dorsal line the dorsal armour was not 

 co-ossified with the ribs, and if co-ossification did occur it could only 

 have happened on the surface of the ilia and along the summits of 

 the neural spines. Through this indication there appears to exist 

 a still greater resemblance between the so-called ' Daniibiosaurus ' 

 bone of Gosau and PolacantJius, because also in that one a median 

 empty space seems to have separated the underlying rib-like bone. 

 As Professor Seeley has likewise mentioned, the iliac bone extends 

 laterally somewhat beyond the lumbosacral shield. 



Besides the dorsolumbar shield Hulke distinguished in his 

 second Polacanthus paper three types of dermal armour. The 

 roof -like plates Hulke supposed to be situated a cheval above 

 and below the haemal and neural spines, while the spines were 

 believed to be situated somewhere on the thoracic region, and flat 

 shields are supposed, according to Fox's interpretation, to have 

 covered the belly. 



Since the dermal covering of no Stegosaurid animal is yet 

 perfectly known, and the figure Marsh gives of Scelidosaurus does 

 not seem quite correct in this regard, it seems at first rather difficult 

 to decide these questions, but having been able to investigate, 

 besides Polacanthus, both Hylceosaurus and Scelidosaurus, and having 

 fitted in Polacanthus numerous new pieces together, I am able to 

 assert that the roof-shaped scutes formed two rows right and left 

 of the neural spines of the caudal vertebra, and that the highest 

 spines belong to the scapular region. 



Before dealing, however, with this problematical armour it is 

 necessary to consider first the nearly perfectly preserved mail of 

 Scelidosaurus. On the body of Scelidosaurus one can trace out quite 

 distinctly seven rows of dermal ossifications, which show that the back 

 of this animal was covered by a median row of keeled scutes, on the sides 

 of which a smaller row is visible. In this second row the tubercles 

 increase the nearer we approach the sacrum. Laterally of this 

 second row we can distinguish on each side two rows of scutes 

 converging forwards, in each of which the scutes gradually augment 

 in height the more they approach the scapular region, so that 

 meeting at the scapula these two rows at last form only one row of 

 rather tall spine-like ossifications. A good idea of the way they 

 augment in height towards the scapula is given in the otherwise not 

 quite correct reconstruction of Scelidosaurus published in the edition 

 of the popular guidebook of the British Museum. 



In the sacral region of Scelidosaurus the dermal ossifications are 



