36 HUXLEY, ON THE SKELETON OF FISHES. 
by Von Siebold and Stannius (1846), the latter (‘ Wirbel- 
thiere,’ p. 10) considers the vertical caudal plate to be - 
produced by the coalescence of the superior and inferior 
arches, interhzemal and interneural bones “ of the posterior 
caudal vertebra or of many of the caudal vertebre ;” and in 
a note it is added, that the commencement of the process 
may be clearly traced in the pike. 
Avaluable paper published bythe late eminent ichthyologist, 
Heckel, in the ‘ Sitzungs-berichte der Kaiserlichen Akademie 
der Wissenschaften’ for 1850 (p. 143 et seq.), contained the 
first accurate and comprehensive account of the structure of 
the piscine tail, and threw quite a new light on the general 
doctrine of the relation between ancient and modern fishes. 
“The few now-living successors of the bony Ganoids with 
complete vertebra, which first appeared in the Jurassic 
period—our Lepidosteus, Polypterus, and probably also Amria 
(the latter of which I have had no opportunity of examining) 
—still have quite imperfect terminal vertebre, behind which 
a part of the chorda persists in a perfectly unossified state. 
At the same time these terminal vertebree appear to be de- 
veloped in quite a different way from those of ordinary 
Teleosteans, for the arrested commencements of the posterior 
caudal vertebre, or their first centres of ossification, appear, 
not as in the latter, above and below at the base of already 
formed spinous processes, but at the sides of the chorda, 
before either spinous processes or vertebral arches are de- 
veloped. They become thickened anteriorly, and penetrate 
like wedges towards the axis of the chorda. Indeed, it would 
seem, from the fact that different imdividuals of these fishes, 
without distinction of size, present a considerable variation 
in the number of their terminal vertebrae (which may be 
even perfectly developed) as if they constantly added new 
vertebree, whereby the end of the-vertebral column—that 1s to 
say, the still naked chorda—must gradually, if not perfectly, 
be converted into ossified bodies of vertebree. .... . 
“Another group of fishes, or rather of the now-living 
Teleosiei (whose origin is wrongly placed in the Cretaceous 
period, since it certainly took place much earlier, in the 
Jurassic epoch), also possess an imperfect vertebral column. 
No inconsiderable portion of the end of the chorda remains 
without developing vertebree throughout the whole life of 
the fish, and becomes hidden under a roof-like arrangement 
of peculiar bones, which, supported upon the penultimate 
vertebral bones and projecting backwards beyond them, and 
seeming to be mere upper spinous processes, or ray bearers, 
unite with the broad inferior spmous processes which haye 
