178 Reviews—Prof. K. v. Zittel’s Palichthyology. 
lateral processes. Sometimes also the hypocentrum consists of two 
pieces, which meet together beneath the notochord. Besides the 
hypocentrum there is seen somewhat above on both sides a keel- 
shaped plate (pleurocentrum), pointed below but rounded above, 
by which the notochord is partly, though not completely, enclosed. 
In the next stage both the upwardly pointed hypocentra and the 
downwardly pointed pleurocentra extend so far, that a zig-zag 
appearance is produced laterally, and as the two pleurocentra also 
often unite on the dorsal aspect, the vertebra is now formed of two 
thin horse-shoe-shaped half-rings, completely encircling the noto- 
chord. In such half-vertebre the upper arched pieces (pleurocentra) 
never partly overlap the lower pieces (hypocentra), as Heckel has 
erroneously stated. In the genus Hurycormus the vertebral column 
consists in its front half of about equally developed horse-shoe-shaped 
half-rings, while in the caudal region the attenuated ends of the 
hypocentra and pleurocentra unite dorsally and ventrally, thus 
forming two closely apposed delicate rings completely surrounding 
the notochord (‘falsche Hohlwirbel,’ Vetter). The caudal region 
of the recent genus Amia exhibits this stage, except that the ossifica- 
tion has proceeded further inwards and to a great extent constricted 
the notochord. In a number of Jurassic Lepidosteoids ( Aspido- 
rhynchus, Belonostomus, Histionotus, Ophiopsis) and Paleozoic 
Crossopterygians (Rhizodopsis, Megalichthys), the vertebral centrum 
consists of a single thin cylindrical ring with a smooth or vertically 
striated outer surface. That the ‘Hohlwirbel’ or ‘ Ringwirbel’ 
results from the lateral fusion of two half-rings, is conclusively 
proved, for example, by Aspidorhynchus, for here the caudal region 
always consists of hollow vertebrae (Hohlwirbel), while the front 
abdominal region mostly consists of half-vertebra, of which the 
hypo- and pleuro-centra form rings closed dorsally and ventrally, 
but remain completely separated by a suture laterally.” 
In the systematic description of the Ganoids, the doubtful frag- 
mentary remains described by Pander from the Upper Silurian of 
the Island of Oesel are briefly noticed; and then follow the three 
problematical “ orders” of Pteraspidee, Cephalaspidee, and Placodermi. 
Alth’s determination of the simple shield named Scaphaspis, as 
placed ventrally beneath the more complex shields of Pteraspis 
and COyathaspis, is adopted; and the restored figure of the two 
connected shields of Pteraspis is reproduced. These fossils are 
common in the Old Red, not of Scotland (as stated), but of Western 
England. We would also remark that Pal@aspis, Claypole, so far 
as defined, falls in the genus Holaspis. A very complete synopsis 
of the Cephalaspide is given, including references to the species ; 
and the histological remarks are illustrated by Huxley’s figures of 
microscopical sections of the shields. Both these groups have been 
so well treated by Ray Lankester, that it is comparatively easy to 
summarize our present knowledge of their various representatives ; 
but with the Placodermi the case is very different, and a carefully 
prepared digest of the existing literature of the subject shows how 
much revision of the typical Huropean genera is required. Such a 
