were unnaturally drawn out vertically. The depth of the 

 body between the soft dorsal and anal fin origins is 63.6 

 mm (97.0% SL). The least depth of the caudal peduncle 

 is 20.1 mm (30.6% SL), this being substantially deeper 

 than in the other two species of the genus. The length of 

 the first dorsal spine is 24.6 mm (37.5% SL). The second 

 dorsal spine cannot be measured, but the third spine is 

 3.9 mm (5.9% SL). The number of dorsal and anal fin 

 rays appears to have been similar, and between about 20 

 and 25. The vertebral column cannot be seen in the ab- 

 dominal region, but the caudal series can be interpreted 

 as either 11, as in Recent species, or 12. The caudal fin 

 appears to have 12 rays. There is a well-developed, es- 

 pecially obliquely placed, supraneural strut supporting 

 the rear end of the carina. The basal pterygiophores have 

 exceptionally well-developed lateral flanges. The form 

 and number of the teeth are poorly indicated as large in- 

 cisors, with at least three probably located in an outer 

 series in the upper jaw. The stout pelvis has a low but 

 distinct dorsal lobe posteriorly, and there is some in- 

 dication that the scales between the posterior end of the 

 pelvis and the anus were enlarged and elongate to sup- 

 port a small fan or dewlap in this region, as in most Re- 

 cent species. The scale pattern is relatively well preserv- 

 ed, and the basal plates appear to bear two or three more 

 or less vertically oriented low emarginate ridges, sub- 

 stantially different from the condition of spinulation in 

 the other two species of the genus. 



In addition to the three species of Balistomorphus, one 

 other species of balistid is known from the Oligocene seas 

 of Europe. Oligobalistes robustus Danil'chenko (1960), 

 known from three specimens of 27 to 60 mm SL from the 

 Maikop deposits of the Caucasus, is the best preserved 

 fossil balistid. It has not been reexamined for this work, 

 and the following is based on the description by 

 Danil'chenko. Oligobalistes differs from Balisto- 

 morphus by having no surface sculpturing on the scale 

 plates, and by the better development of the second and 

 third dorsal spines (respectively 15 and 10% SL in 

 Oligobalistes versus 8 and 4 to 6% SL in Balisto- 

 morphus), although the first dorsal spine of Oligobalistes 

 is slightly less well developed (30% SL) than in one of the 

 species of Balistomorphus (B. orbiculatus, 37.5% SL). 

 DanH'chenko believed Oligobalistes to differ from all of 

 the Recent genera by the large size of the first dorsal 

 spine and the presence of longitudinal rows of spinules 

 along its anterior edge. The first dorsal spine is slightly 

 longer, but not stouter, than in the adults of any Recent 

 genera, some of which (e.g., Rhinecanthus) have 

 longitudinal rows of spinules along the anterior and 

 lateral edges of the scales that are as well developed as in 

 Oligobalistes. If the scales of Oligobalistes are com- 

 pletely smooth then it also differs in this respect from the 

 Recent genera, all of which have at least low emar- 

 ginations or spinules. 



The teeth in Oligobalistes were described as num- 

 bering four to six in each jaw, but presumedly any num- 

 ber above the four in an outer series (as found in all Re- 

 cent species) represent misplaced teeth from another 

 outer series or from any of the inner series. The vertebral 



column was described as 7 -I- 10, but the illustration 

 clearly shows 11 caudal vertebrae, and a 7 -I- 11 column, 

 as in all Recent species. 



Danil'chenko thought Oligobalistes most closely 

 related to the Recent Balistes, but because of the ab- 

 sence of a preocular groove, Oligobalistes is probably 

 more closely related to Balistapus and Rhinecanthus, as 

 discussed under Generic Relationships. 



Generic relationships.— The generic relationships 

 within the Balistidae are not especially clear on the basis 

 of the present work. Features that a priori seem general- 

 ized on the basis of the evolution of balistids from tria- 

 canthids are a terminal mouth, well -developed third dor- 

 sal spine and overlapping scale plates. 



The relatively straight-edged or gently curved heavy 

 crushing incisors of triacanthids contrast to the slightly 

 fewer in number, but at least superficially larger, notch- 

 ed teeth basic to balistids. While slightly larger in sur- 

 face area, balistid teeth are thinner edged and less heavy 

 than in triacanthids, except in very large individuals of 

 most species of balistids, in which the teeth become just 

 as massive if not more so than in triacanthids, while still 

 retaining at least some degree of notching. The presence 

 of unnotched edges on the more medial teeth in adult 

 Melichthys is not an indication of a generalized con- 

 dition, for Berry and Baldwin (1966:449) have shown that 

 the teeth are normally cusped in juveniles and only 

 gradually do the more medial teeth become worn down to 

 a relatively straight edge in adults, definite remnants of 

 the notching always being present on the more lateral 

 teeth. 



However, the dentition of Xanthichthys and Odonus 

 can surely be considered a specialization, with the en- 

 largement of the second most medial tooth in the 

 premaxillary and, in Odonus, of its cusp, into a fang. The 

 second most medial tooth in the dentary is also often 

 slightly larger in Xanthichthys and Odonus than it 

 would be in other balistids. In contrast to all other 

 balistids, which have white teeth, the teeth oi Xanthich- 

 thys have a dark tinge, at least as adults (Berry and 

 Baldwin 1966:456), and those of Odonus are a deep red. 

 Also in contrast to other balistids, the teeth in Xanthich- 

 thys and Odonus, as well as in Canthidermis, the only 

 truly oceanic balistid, do not become exceedingly mas- 

 sive and thickened in large adults, the dietary changes 

 between juveniles and large adults in these three genera 

 apparently not being at all as extensive as in other 



The specialized condition of the anterior displace- 

 ment of the suspensorium and lower jaw is found most 

 prominently in Xanthichthys and, especially, Odonus, 

 the mouth being distinctly supraterminal. The mouth in 

 Melichthys is slightly supraterminal, but in other 

 balistid genera it is essentially terminal. 



Probably not much importance should be placed on 

 the size of the third dorsal spine as an indicator of 

 specialization and relationship. The third spine is not in- 

 volved in the locking mechanism of the first two spines, 

 and it has neither a complex basal articulation 



120 



