449 
seen by Ransom and Tmompson (’86) and Junin (°87) take place is 
explained by Kuprrer (’96) in the following way: “Es lassen sich 
bei ganz jungen Larven bis zu 4 mm Länge Zellengruppen in der 
Nähe der Chorda nachweisen, die mit Spinalnerven Verbindungen haben 
und die ich für sympathische Ganglienanlagen zu halten geneigt war. 
Ich konnte aber die weiteren Schicksale dieser tiefen Zellenhaufen 
nicht verfolgen. Spätere Stadien gewähren hier auch an den besten 
Präparaten keinen Einblick. Pigment, compactes Bindegewebe, dann 
Knorpelgerüste an der Chorda verdecken diese Bildungen. Es kann 
aber wohl nicht bezweifelt werden, daß aus diesen Zellengruppen ein 
bleibender mit den Spinalnerven zusammenhängender Nerv hervorgeht, 
der sich an dem einwärts verlagerten Branchiogastricus anlagert und 
später mit demselben in eine Scheide eingeschlossen wird. Auf diese 
Weise allein könnte ich mir das Zustandekommen dieser Verbindung 
erklären.” 
The facts which I have obtained confirm the opinion expressed 
by Kurrrer (96) that the so-called ramus recurrens vagi is 
composed of the fibres of ventral spinal nerves. The method of their 
union with the “Epibranchialstrang” of the vagus as a bundle of thick 
fibres lying above the smaller fibres of this nerve may, I believe, be 
made clear by a comparison of two stages of development. In an 
embryo of 8 mm (20 days) the ventral motor root of the 12th myo- 
tome passes ventrad to innervate the ventral portion of this myotome 
behind the ‘ramus recurrens vagi” without uniting with it. In an 
embryo of 5 cm, as the result of the backward extension of the branchial 
basket carrying with it the epibranchial portion of the vagus, the 
ventral root of the 12th somite passes ventrad as a component part 
of the “ramus recurrens vagi”. In this way in earlier stages 
the ventral roots of myotomes 7—11 inclusive which were primarily 
situated behind the last visceral arch, have united one by one as the 
branchial basket elongates !). In a precisely similar way the ventral 
spinal roots which form the primitive hypoglossus nerve in Squalus 
unite with one another and (for a short distance) with the epibranchial 
portion of the vagus, before they pass behind the last visceral cleft 
to innervate that portion of the ventral longitudinal muscle which lies 
anterior to the coracoid. From these facts, I conclude that the so- 
called ramus recurrens vagi is homologous with the 
hypoglossus nerve of higher Vertebrates. In how far the 
1) A similar explanation of the course of the hypoglossus nerve in 
Mammals has been given by Frontier (’85). 
