tyO CO basibranchials 

 l_J' —hypobranchials 



branchiostegal 



ventral hypohyal 

 ceratohyal 



Figure ill .—Lactoria comuta: dorsal view of branch- 

 ial arches (extended on lower side); lateral view 

 of hyoid arch and urohyal; 88.2 mm SL, Philippines. 



present in all species, with the exception that only a 

 minority of the specimens of Acanthostracion quad- 

 ricornis examined had a first pharyngobranchial present, 

 at least as an ossification, and that the first pharyngo- 

 branchial may also have been absent in the single 

 specimen of A. notacanthus examined, the specimen 

 having been poorly cleared and crudely dissected in this 

 region and the presence or absence of that pharyngo- 

 branchial now being impossible to decipher. In most of 

 the species examined the third pharyngobranchial is 

 toothless, but minute teeth are present on it in Lac- 

 tophrys trigonus, Tetrosomus concatenatus, and in most 

 but not all of the specimens examined of Acantho- 

 stracion quadricornis and Lactoria cornuta. Such varia- 

 tion is to be expected in what appear to be highly 

 rudimentary structures, and if more specimens of the 

 species in which no third pharyngobranchial teeth were 

 found had been examined, the suspicion is that at least 

 some of these species could be added to the list above of 

 those known to at least sometimes have teeth on this ele- 

 ment. 



Generic relationships and comparative diagnoses 

 of subfamilies (Ostraciinae, Lactophrysinae).— Fra- 



ser-Brunner (1941c) divided the ostraciids into two sub- 

 families, the Indo-Pacific Ostraciinae and Atlantic Lac- 

 tophrysinae, on the basis of the lower number of dorsal, 

 anal, and pectoral fin rays in the former and of supposed 

 vertebral differences between the two groups. Tyler 

 (1963a:185) showed that the vertebral differences men- 

 tioned by Fraser-Brunner did not stand close examina- 

 tion of a larger number of species than Fraser-Brunner 

 apparently had examined, but Tyler only hinted at a 

 few other vertebral characters that might eventually be 

 used to separate the two groups. These and additional 

 characters that distinguish the two subfamilies are dis- 

 cussed more fully in the preceding section on anatomical 

 diversity in the family. 



The separation of the ostraciids into two phyletic lines 

 of subfamilial rank seems justified to me on the basis of 

 the following differences: 



OSTRACIINAE 



dorsal and anal fin rays modally 9 

 dorsal and anal fin basal pterygio- 



phores modally 8 

 pectoral fin rays modally 10 



four vertebrae involved in the fu- 

 sion complex at the rear of the 

 skull, except for one species with 

 five 



two or more postdorsal vertebrae 

 with trifid neural spines 



last anal fin basal pterygiophore 

 laterally or anterolaterally ex- 

 panded into a prominent fiange 

 for carapace support 



last dorsal fin basal pterygiophore 

 moderately or greatly expanded 

 laterally and posteriorly for cara- 

 pace support 



haemal spine of penultimate verte- 

 bra fused to the centrum, either 

 with a foramen or solid 



myodome small and shallow, but 

 present 



Indo-Pacific distribution. 



LACTOPHRYSINAE 



dorsal and anal fin rays modally 10 



dorsal and anal fin basal pterygio- 

 phores modally 9 



pectoral fin rays modally 12, ex- 

 cept one species with 11 



five vertebrae involved in the fu- 

 sion complex at the rear of the 

 sliull 



all vertebrae with unbranched 

 neural spines 



last tinal fin basal pterygiophore 

 not expanded for carapace sup- 

 port 



last dorsal fin basal pterygiophore 

 not expanded for carapace sup- 

 port 



haemal spine of penultimate verte- 

 bra autogenous or fused to the 

 centrum, always with a foramen 



myodome essentially absent 



Atlantic distribution (to South 

 Africa). 



If, as suggested, the two subfamilies represent separate 

 phyletic lines, it is assumed that the involvement of the 

 fifth vertebra in the anterior fusion complex of the Indo- 

 Pacific Lactoria fornasinii has taken place indepen- 

 dently by convergence from that in the Atlantic species, 

 and that the fusion of the haemal spine of the 

 penultimate vertebra to its centrum in the four species of 

 Atlantic Acanthostracion has been similarly indepen- 

 dent from that in the Indo-Pacific species. This, it seems 

 to me, is far easier to believe than that the entirely con- 

 stant differences: 1) the numbers of dorsal, anal, and 

 pectoral fin rays; 2) the structure of the neural spines of 

 two or more postdorsal vertebrae; 3) the structure of the 

 last basal pterygiophore of the dorsal and anal fins; and 

 4) the different distributions and myodome develop- 

 ment, between the two groups have no phylogenetic 

 meaning. 



