316 DK. C. W. AADREWS ON A XEW SPECIKS OF 



projecting forwards and backwards so as to unite with one 

 another in a short suture ; external to this they narrow rapidly 

 to tlie edge of the broad and strongly developed rib. Towards 

 the middle line of the carapace the costals are free from the ribs 

 for a short distance, forming flanges which project towards the 

 vestigial neurals, but do not reach them. In this respect the 

 degree of reduction is intei'mediate between what is seen in 

 Protostega and Archelon,'n\ which the greatly reduced costals still 

 meet the neurals and unite with one another, and in Dermochelys, 

 where the upper borders of" the costals, though marked by a 

 distinct ridge, do not form flanges projecting towards the 

 neui-als, which are, in fact, absent oi- perhaps in part representeil 

 by the bilobate upper ends of the neural spines; moreover, in 

 DermocJtelys the costals do not unite with one another. The 

 upper ends of the ribs are thick and no doubt had a considerable 

 articulation with the vertebral centra, l)ut this region is abraded 

 and only the portion articulating with the neural arch remains. 

 The distance between the heads of the ribs of opposite sides is 

 about 33 mm. The most anterior of the ribs preserved appears 

 to be curved backwards towards the rib behind and to have had 

 the costal plate on the hinder side only: this was, probably, the 

 first rib. The outer free portion of the other ribs is very broad 

 and strongly developed. 



The neural arch beai'S a short stout neural spine, to the upper 

 end of which is attached a thin, flat, table-like neural, which 

 projects on either side for about 7 mm., Ijut remains separated 

 from the upper edge of the costal by an interval of 10 mm. 



In the figure (PI. II. fig. 4, text-fig. 2) the relations of the 

 costals, neurals, ribs, and neural arches to one another and to the 

 epithecal shell is well shown. It will be noticed that the median 

 ridge of the eipithecal shell is not immediately above the middle 

 line of the underlying neural, but in the crushing that has been 

 undergone has been displaced to one side. This probably indi- 

 cates tha,t in life the epithecal and vestigial thecal shells were 

 separated by a considerable layer of soft tissiie now represented 

 by a thin film of matrix only. 



Of the i-emainder of the skeleton nothing is yet known. It is 

 greatly to be desired that engineei's and others who are on 

 the spot where excavations are opened, would collect all the 

 fragments that are exposed, especially in tropical localities like 

 ISTigeria, where fossils once uncovered are rapidly disintegrated 

 and cuttings very soon overgrown and rendered inaccessible. Of 

 course, it would be still bettei' if, when such works are in pi'ogress, 

 a skilled collector could be present to see that nothing was lost. 



The Chelonian above described approaches most nearly to 

 Ps&phojjhorus^ but, as pointed out, differs from this in sevei'al 

 respects, notably in the arrangement of the plates on the 

 epithecal shell, and it is thei'efoie referied to as a new genus 

 Cosmochelys, the specific name being Vosmochelys dolloi in honour 



