N. ROSEN, STUDIES ON THE PLECTOQNATHS. 5 



ventral arches. This author vvrites, p. 71: »Elles naissent 

 de la colonne vertebrale par deux racines, dont Tiine s'at- 

 tache au corps de la vertébre, et Tautre a la base de Tapo- 

 physe épineuse; puis elles divergent en formant, des deux 

 cotés de la colonne vertebrale, des petitets éminences trian- 

 gulaires.» According to Dareste the ribs are attached to 

 the vertebrae, except the two anterior paires, which attaché 

 at the i>haemapophyses>. Regan characterizes the vertebral 

 column of the Triodontids as follows, p. 288: »praecaudal 

 vertebrae without parapophyses; epipleurals well-developed.» 

 The question concerning the structure and nature of the trans- 

 verse processes and the ribs of the Triodontids must be left 

 undecided until genetical researches have been made. 



Tetrodontidae. 



(Fig. B; Pl. 2, Fig. 1-2.) 



I have studied Spheroides iestudineus, both adult speci- 

 mens and young ones measuring 18 mm. in length. The dor- 

 sal arches of the three anterior vertebrae do not touch each 

 other in the median line and thus a »spina bifida» is for- 

 med. The neural canal, however, is not open dorsally but 

 bordered by a ligament (Fig. B II). In this ligament paired 

 bones occur. They are laterally in connection with the dor- 

 sal arches and form in adult specimens an almost completely 

 ossified roof of the neural canal. To judge from sections of 

 young specimens the skeleton pieces in question have proba- 

 bly been, at least partly praeformed by cartilage. They cor- 

 respond without doubt to similar cartilaginous pieces found 

 in the same ligament in other teleosts and described by se- 

 veral writers. The views put forth about their nature are 

 different. Scheel considers, p. 22, that such pieces »den 

 eigentlichen dorsalen Abschluss der Neurapophysen darstellt», 

 while Shauinsland expresses his opinion in the following 

 words, p. 468: >Sie werden wahrscheinlich, ebenso wie bei 

 Lepidosteus, als dorsale Reste von Bogen des kranialen Skle- 

 rotomstiickes zu betrachten sein, deren basalen, der Chorda 

 aufsitzenden Teile knorpelig entweder gar nicht mehr, öder 

 jedenfalls von der Basis des eigentlichen Bogens nicht mehr 

 getrennt, zur Ausbildung gekommen sind.» The structure of 



