KIMMERIDGIAN DINOSAURS. 565 



Each neurapophysis at its confluence with tlie ceiitnmi gives a triangular horizontal 

 section (PI. 67, fig. 3, np)^ the base of the triangle, 1 inch 5 lines, being anterior, the 

 obtuse apex behind. The inner, shorter side, next the neural canal, is parallel with its 

 fellow and the trunk's axis, the outer side, 2 inches 9 lines in extent, slopes from the 

 broad fore part backward and mesiad to the hind margin of the neural arch. 



From the upper and anterior forwardly sloping part of each neurapophysis the 

 prezygapophysis (~) is developed ; it is short, thick, obtuse, with a flat articular surface, 

 looking upward, inward, and slightly forward ; subcircular, an inch in diameter. From 

 the narrower hind part of the neural arch the common base of the pair of postzyg- 

 apophyses {:', ^') rises, expanding to form their articular surfaces, which look in directions 

 opposite to those in front. The hind surface of the common base of these articular expan- 

 sions has a wide and deep vertical channel.* 



The neural spine (««) is subquadrate at its base, with the lateral angles broadly rounded 

 off (PI. G8, fig. 2, ns)- The line of attachment of the base of the spine rises from before 

 backward (ib., fig. 3). A median anterior ridge (PI. 67, fig. 1, x) strengthens the lower 

 half of that surface, as a similar but thicker ridge (PI. 68, figs. 1 and 3, «) does the 

 posterior corresponding tract. Where these ridges cease the spine begins to expand into 

 its rough obtuse summit, chiefly transversely, so as to give it an elliptical contour 

 extended in that direction (PI. 67, fig. 2). 



The foremost of the caudal vertebrae remains in the block of matrix with the sacrum. 

 The present I take to be the second of the series. There is no trace of hypapophysis for 

 a hsenial arch in either of these caudals (the under surface of the centrum of the second is 

 figured in PI. 67, fig. 4). In Scelidosaurus the first or foremost caudal alone is devoid 

 of haemal arch ; in the second caudal the lower part of the hind border is touched by 

 the smaller anterior facet on the base of the haemapophysis. 



In the few succeeding caudal vertebrae, with diminution of general size, the vertical 

 extent and the length of the plenrapophyses decrease in a greater ratio. A larger 

 proportion of the side of the centrum is left free below the rib's confluence therewith ; 

 and this free surface of the centrum shows, as in the specimen selected for PI. 69, an 

 upper C^) and a lower (<■') depression. The transverse extent of the centrum decreases 

 without corresponding loss of vertical extent. The hind surface of the centrum (ib., fig. 2) 

 becomes more concave, without corresponding increase of convexity of the fore surface. 

 The contour of the hind surface approaches the subhexagonal. 



The anterior and posterior ridges of the neural spine subside ; the fore ridge is longest 

 retained, but shrinks toward the base of the spine, as at r, fig. 1. In the subject of this 

 Plate, as in three other caudals extracted from the matrix, the neural spine has been bent to 

 one side, as shown in PI. 69, fig. 2. This distortion I conceive to be due to movements of 

 the matrix after the fossil had been inclosed thereby and become petrified therewith. For, 



* It is possible that a similar facet may have been ligamentously attached to tiie rough surface 

 exteuded from tlie lower margin of the terminal surface. 



