ELASMOBRANCHII. 337 



" spiracles." By means of these water can be admitted to the 

 pharynx, and thence to the gills. 



By Professor Owen the Plagiostomi are divided into three 

 sections, termed respectively the Cestraphori, the Selachii, and 

 the Batides. 



a. Ccstraphori. In this division there is a strong spine in 

 front of each dorsal fin, and the back teeth are obtuse. The 

 only living representative of this group is the Port Jackson 

 Shark (Cestradon Philippi), characterised by its pavement of 

 plate-like crushing teeth, adapted for comminuting small Mol- 

 luscs and Crustaceans. It is exclusively an inhabitant of the 

 Australian and Chinese seas, and is remarkable for its close re- 

 semblance to a large group of extinct forms, of which the best 

 known are the genera Hybodus and Acrodus from the Second- 

 ary Rocks. 



The Cestraphori are known in a fossil state mainly by their 

 fin-spines, or " ichthyodorulites," and their teeth. It is obvious, 

 however, that it must be often very difficult or altogether impos- 

 sible to determine absolutely whether a spine or a piece of shag- 

 reen belongs to a Cestraciont or to a true Shark. Some of the 

 forms, therefore, to be immediately mentioned, must be regarded 

 as being only provisionally placed amongst the Cestraciontida. 



With this proviso, the earliest known traces of Cestraciont 

 fishes appear to be in the Upper Ludlow Rocks, at the sum- 

 mit of the Silurian series. Here, within the limits of a single 

 stratum, well known as the " bone-bed," occur remains which 

 have been with more or less probability referred to Cestra- 

 cionts. Some of these (fig. 295) are in the form of compressed, 



Fig. 295. A, Spine of Onchus tenuistriatus. B, Shagreen-scales of Thelodus. 

 Both from the bone-bed of the Upper Ludlow Rocks. 



slightly curved spines, the sides of which are grooved longi- 

 tudinally. These have been referred to a provisional genus, 

 under the name of Onchus, and there appears to be little doubt 

 as to their truly belonging to fishes of some kind. It is, how- 

 ever, quite possible that they really belong to Pteraspis, in 

 which case they must be removed from their present place. 



Along with the spines of Onchus arb found fragments of 



prickly skin or shagreen, which have been referred to the 



temporary genus Sphagodus, along with minute cushion-shaped 



bodies, which are doubtless placoid scales, and which have 



Y 



