292 MR. C. TATE REGAN ON [Nov, 4, 



It has been asserted that the term spina bifida is not correctly 

 applied to the neural spine of the anterior vertebrae in these 

 fishes, because the neural canal is closed ; but in the Molidfe this 

 is not the case, and in the Diodontidte the neural canal is open 

 above in the posterior prascaudal region ; the anterior bifid spines 

 are in all cases obviously homologous and forming one series with 

 the single neural spines which succeed them, and when the 

 neural canal is closed by a bony roof this must be regarded as a 

 secondary feature, due to the meeting of outgrowths from the 

 base of the neural spine of each side after they have separated. 



Family 1. Tetrodontid^. 



Praecaudal vertebra? without parapophyses, the first four or five 

 with bifid neural spine and closed neural arch ; no epipleurals. 

 Prseorbital not ossified ; palatine firmly united to the skull ; no 

 distinct bony nasal cavity ; premaxillaries not protractile, united to 

 maxillaries ; teeth in the jaws coalescent, in each forming a beak 

 with median suture ; palate toothless ; fourth vipper pharyngeals 

 present, toothed ; lower phaiyngeals sepaiute ; interoperculum a 

 long rod, attached to inner face of prseoperculum, sometimes 

 connected with operculum, never with suboperculum. Nostrils 

 various. Four branchial arches, the fourth not bearing a gill, 

 not followed by a slit ; pseudobranchise pi'esent ; six branchio- 

 stegals, the first a broad plate. Skin naked, usually with movable 

 spines, rarely with bony plates. Caudal peduncle normal. 

 Skeleton well-ossified. Belly very inflatable. Air-bladder present. 



Many authors have failed to understand the evolution of the 

 nasal organs in this family, as is shown by the wording of their 

 diagnoses, such phrases as "nostrils represented by two solid 

 tentacles on each side," " nostril with a tube," &c. being quite mis- 

 leading. In the more primitive forms {Lagocephalus) there are 

 two nostrils on each side, situated in an oval nasal area, which 

 overlies an internal nasal sac, exactly as in Balistes, Triacanthus, 

 &c. From these we pass to fishes {Spheroides) in which the nasal 

 area is raised up into a more or less prominent tubular papilla 

 bearing the two nostrils, whilst the nasal sac is scarcely sunk below 

 the level of the skin, and is in great part represented by the interior 

 of the papilla, on the walls of which are the terminations of the 

 olfactory nerve. By the absorption of the septum between 

 the nostrils at the end of the papilla they become confluent, and 

 we get a circular tube produced terminally into two more or less 

 distinct lips or tentacles, in the more specialized of which the 

 circular tube is short and constricted, so that we have two 

 tentacles, on the inner surface of which are the terminations of 

 the olfactory nerve, united basally. Thus when the nostrils 

 become confluent the interior of the nasal sac is exposed, and in 

 some species of Tetrodon it may be said to be raised above the 

 level of the skin. In Tropidichthys the circular tube has de- 

 generated to an inconspicuous rim vdi\\ a minute aperture. In 



