1885. ] Embryology. 9I 
cal fishes, succeed each other in the order of time. A sixth ex- 
ceptional form is developed in consequence of an extensive 
degeneration of the chordal axis and hinder end of the urosome, 
unaccompanied by an upbending of the hinder end of the axis, 
as in the case of the evolution of heterocercy. The most prim- 
æval stages, or those found to appear in the younger phases of 
the growth of fishes are somewhat approximated by the structure 
of the fins of some of the most ancient Devonian, Triassic and 
Jurassic forms and by such living forms as Chimæra, the Dipno- 
ans and Leptocardians, but the parallelism of the development of 
the tail of young fishes with the successive modifications of caudal 
structure found in the forms of ‘successive geological periods is 
not.exact, as we shall presently show. 
1. Archicercy—The most primitive modification of the urosome 
is that which I will call avchicercal, and which is without any 
median fin-folds whatsoever. While it is true that only a few de- 
generate or specialized forms of true fishes (Hippocampus, Nero- 
phis) approximate such a condition, it must be admitted that the 
fins are acquired structures, and that the folds from which they 
are developed have been acquired in the course of the evolution 
of the ancestry of the fishes. When a young fish is developing in 
the egg its tail grows out at first as a blunt prolongation back- 
wards, which is for atime wholly without fin-folds, cylindrical 
and vermiform in general appearance, with the muscular somites 
clearly marked. - 
The larva of Branchiostoma (Fig. 1) is at first without median 
fin-folds, and that of Petromyzon seems to be without them during 
the very early stages, and while we must make due allowance in 
both these cases for the effects of degeneration, we may, I think 
it probable, look upon these types as possessing at one stage a 
typically archicercal and vermiform tail. The solitary Urochorda 
or Ascidians pass through an archicercal stage of development 
of the urosome. In the course of further development the As- 
cidians never seem to pass beyond what I have called the second 
or lophocercal stage when it is absorbed in the caducichordate 
forms, but persists in the same stage in the perennichordate Ap- 
pendicularia, 
. The Elasmobranchs seem to pass through an archicercal stage 
while the Amphibians do not exhibit it in so pronounced a way, 
very soon becoming lophocercal, though the larva of Dactylethra 
has the anterior part of the urosome with high median fin-folds 
while the termination is somewhat like that of C/iimera monstrosa 
(Fig. 2), but tapers more and is typically archicercal (teste, W. K. 
