^L\'<^i.j^.:'^i 



tetraodontids studied in lacking even a trace of a mesop- 

 terygoid. 



On the basis of the material examined, I am unable to 

 distinguish Fugu from Torquigener, and I suspect that 

 when more species of Torquigener are examined the dis- 

 tinction would be even more difficult to make. Ambly- 

 rhynchotes is obviously closely related to Torquigener 

 and Fugu, differing from them by the tendency to have 

 the frontal evenly narrowed anteriorly and to have 

 prickles on the side of the body. Much more remains to 

 be done on the external and internal morphology of these 

 Indo-Pacific genera with two nostrils and, most usually, 

 two lateral lines. 



It is pure surmise, but I would guess that an early 

 Sp/ioeroides-like stock with a generalized moderately 

 wide skull and evenly tapered frontals in the Indo-Pacific 

 gave rise to Amblyrhynchotes, Torquigener, and Fugu, 

 with the major changes from the ancestral type being the 

 usual development of a lower lateral line, the usual shor- 

 tening of the ethmoid-vomerine region, the constant 

 development of the dorsal flange of the parasphenoid, 

 and the frequent anterodorsal prolongation of the sphen- 

 otics. If this hypothesis is correct, the ancestral Sphoe- 

 roides-like group has subsequently become extinct in the 

 shallow waters of the Indo-Pacific, and is represented 

 there only by a single deepwater species. 



All of the genera discussed above with two nostrils (the 

 "Lagocephalidae" and "Colomesidae") were said by 



Figure 269.-Dorsal views of skulls of: left, 

 Arothron atellatus, ca. 420 mm SL, Seychelles; 

 right, A. armilla, 61.3 mm SL, Australia. 



Fraser-Brunner (1943) to have the prefrontals separated 

 on the dorsal surface of the skull mainly by the frontals, 

 while in Ephippion, Arothron, Monotreta, Chelonodon, 

 Carinotetraodon, and Tetraodon (the "Tetraodonti- 

 dae"), all with a single nostril, the prefrontals were said 

 to be separated mainly by the ethmoid (and the sphen- 

 otics separated from the prefrontals by the frontals, as in 

 the "Lagocephalidae"). 



Whether one considers the ethmoid or the frontal to be 

 the main element separating the prefrontals is often a 

 highly subjective decision. In only a minority of species 

 of both groups can it be said clearly that the prefrontals 

 on the dorsal surface of the skull are nearly exclusively 

 separated by either the frontals or the ethmoid. In the 

 great majority of species of both groups the ethmoid ex- 

 tends posteriorly approximately to the level of the pos- 

 terior end of at least the main body of the prefrontals, 

 although much of this posterior region of the ethmoid is 

 often overlaid by the frontals and thus cannot be seen in 

 dorsal view. In Sphoeroides, Lagocephalus, and Colome- 

 sus, for example, the prefrontals are separated about half 

 by the ethmoid and half by the frontals, and there is 

 much variation between species. In Amblyrhynchotes 



332 



