1885. | Embryology. = 
the permanent caudal. It subsequently degenerates, or it may 
persist as a prolongation of the chordal axis covered by integu- 
ment, as in Chimera monstrosa (Fig. 2) or, as in heterocercal 
Amiurus (Fig. 8), it may, at an early stage, have the chorda ex- 
serted beyond the last hypural car- , Es 
tilages and at some distance behind “sy/ 
them have another hypaxial car- / CANNE 
tilage (of) developed, which may be 
called opisthural, as it probably rep- 
resents the remnant of proximal 
hypural pieces, which were devel- 
oped in some more primitive ances- 
tral form in which diphycercy was 
more pronounced or even perfect. 
Where the caudal, ray-bearing fin- 
fold is developed nearer the end of 
the chordal axis (Apeltes, Siphosto- 
ma, Gambusia,) heterocercy is not 
so pronounced, as the urostyle is shorter and only one or two 
of the terminal vertebre are involved, whereas in other cases 
(Salmo, Lepidosteus) more terminal vertebrze may be implicated 
by degeneration. In archaic forms of heterocercy there may be 
epaxial rays and intermediary supports developed, while the hyp- 
axial supports and rays extend to the end of the upwardly bent 
termination of the axial column (Fig. 5). This trait may possi- 
bly differentiate the archaic type of heterocercy (Palzoniscus, 
Platysomus, Acipenser, Squali) from the more recent or special- 
ized form (Amiurus, Fig. 8) now prevalent amongst Teleosts, and 
which have for the most part a more or less well-developed uro- 
style, but with a very short or included opisthure (= dorsal lobe, 
A. Agassiz), and with the epaxial spines of the urostyle displaced, 
rudimentary or aborted. Outwardly homocercal Palzeozoic fishes 
(Dapedius, Pycnodus,) probably had an opisthural filaments de- 
veloped during their larval stages which subsequently became 
aborted, as in Lepidosteus, but in others (Platysomus, Pygopterus,) 
the terminal part of the chordal axis doubtless became segmented, 
the segments bearing hypaxial caudal rays and few or no epaxial 
ones, so that their opisthures were probably rudimentary or 
wanting. 
It thus becomes evident that the development of modern Tel- 
eosts presents only a partial or inexact parallelism with that of the 
Palæozoic Rhomboganoidei, for few, if any, of these forms show 
the urostyle so distinctly developed or the hypural pieces so ex- 
tensively codssified as in existing Teleostei, and we have also 
shown that there is no such thing even as an exact parallelism to 
be discovered between the development of the tail of the embryos 
of the latter and that of the embryos of an existing representa- 
tive of Palaozoic forms, viz., Lepidosteus (Fig. 6). The Rhom- 
