eemed to strengthen this view.

Following Ehlers,[413] Eisig found the homologue of the notochord in the
accessory intestine of the _Capitellidae_ and _Eunicidae_, which he
supposed might easily be transformed, according to the principle of
function-change, from a respiratory to a supporting organ. He finally
disposed of the alternative notion that the notochord was represented in
Annelids by the "giant-fibres" or neurochordal strands which lie close
above the nerve-cord, a view held by Kowalevsky,[414] and for a time by
Semper. These strands were shown by Eisig, and by Spengel, to be the
neurilemmar sheaths of thick nerve fibres which had in many cases
degenerated. The view that the content of the neurochordal tubes was
nervous in nature was first promulgated by Leydig in 1864.

Much difference of opinion reigned as to the true homologies of the
brain and mouth of Annelids and Vertebrates. Beard[415] and others got
over the difficulty of the haemal position of the cerebral ganglion in
Annelids by supposing that it degenerated and disappeared altogether in
the Annelidan ancestor of Vertebrates, and that accordingly it had no
homologue in the Vertebrate nervous system. Beard put forward also the
ingenious theory that the hypophysis represents the old Annelidan mouth.

Van Beneden and Julin[416] assumed that in the ancestors of Vertebrates
the oesophagus shifted forward between the still unconnected lobes of
the brain to open on the haemal surface.

The fundamental assumption of the Annelid theory, that dorsal and
ventral surfaces are morphologically interchangeable, seemed rather bold
to many zoologists, and Gegenbaur[417] voiced a common opinion when he
rejected as unscientific the comparison of the ventral nerve cord of
Articulates with the dorsal nervous system of Vertebrates.

The _Balanoglossus_ theory of Vertebrate descent also belongs, at least
in its first form, to the earlier group of evolutionary speculations.
The gill-slits of _Balanoglossus_ were discovered by Kowalevsky as early
as 1866.[418] _Tornaria_ was discovered by J. Mueller in 1850, but by him
considered an Asterid larva; its true nature as the larva of
_Balanoglossus_ was made out by Metschnikoff in 1870, who also remarked
upon its extraordinary likeness to the larvae of Echinoderms.[419] That it
had some relationship with Vertebrates was recognised by Semper,
Gegenbaur and others, but the full working-out of its Vertebrate
affinities is due to