basibranchials 



:^^^ ' ' ^'" ' ' "^ -"^^v' " hypobranchials 



^ - ceratobranchials 



urohyal 



-c:2ss. 



S. f-^ 



branchiostegal rays 



The nasal sac beneath the surface usually has a single 

 longitudinal upraised fold or lamella. 



Generic relationships. — Given their evolution from 

 balistids, one would expect generalized monacanthids to 

 have a relatively well-developed pelvis with a dorsal lobe 

 and at least rudiments of the fin-ray element hidden 

 within a flexible series of encasing scales of a maximum 

 number of segments. Likewise, the basal pterygiophore 

 supporting the dorsal fin spines would be expected to be 

 well developed and over the rear part of the skull (i.e., 

 the least distance migrated forward) and with the skull 

 the least flattened for articulation with it (i.e., with some 

 remnant of the medial supraoccipital crest of balistids). 



Other generalized features to be expected a priori 

 would be an only slightly elongate body and terminal 

 mouth, six branchiostegals, a groove posterior to the dor- 

 sal spines into which they are folded more or less flush 

 with the surface when unerected, the least reduction in 

 the number of teeth, the least increase in the number of 

 vertebrae, and the presence of an upper free hypural. 



Of the above features, I tend to place greatest em- 

 phasis on the pelvic apparatus and the form of the top of 

 the skull in relation to the basal pterygiophore of the 

 spiny dorsal fin. The supraoccipital is best developed in 

 Stephanolepis, Monacanthus, and Paramonacanthus, 

 reaching to the front of the eye, and only in these three 

 genera does the supraoccipital retain a vertical medial 

 crest anteriorly for articulation with the laterally com- 

 pressed platelike anterior region of the spiny dorsal fin 

 basal pterygiophore. I take this low anteriorly placed 

 supraoccipital crest as the remnant of the much sturdier 



crest that is found more or less throughout the length of 

 the balistid supraoccipital and which ends posteriorly in 

 the thick lateral expansion which helps, along with the 

 epiotics, to support the anterior end of the first basal 

 pterygiophore of the spiny dorsal fin. 



Stephanolepis, Monacanthus, and Paramonacanthus 

 also have the pelvic apparatus well developed, with the 

 dorsal lobe of the pelvis of moderate to large size and 

 with the maximum monacanthid number of three seg- 

 ments of flexible encasing scales surrounding a hidden 

 fin-ray element to which muscles attach through ten- 

 dons running along the dorsal and ventral surface of the 

 pelvis, the dorsal tendon coursing through a canal in the 

 base of the dorsal lobe. In all three of these genera the 

 dorsal spines are placed relatively posteriorly on the 

 skull, at a level behind the middle of the eye. 



In most other genera of monacanthids examined (Per- 

 vagor, Amanses, Cantherhlnes, Acanthaluteres, Oxy- 

 monacanthus, Rudarius, Laputa, Navodon, Pseuda- 

 iuteres, Paraluteres, Psilocephalus) the supraoccipital 

 does not extend as far forward as in Stephanolepis, 

 Monacanthus, and Paramonacanthus and there is no 

 crest at all. With the exception of Laputa (and Chaeto- 

 derma as well), the former group of genera are also 

 specialized in having an inflexible series of encasing 

 scales, or none at all, and no fin-ray element, other than 

 sometimes a mere cartilaginous plug at the end of the 

 pelvis. 



Laputa is the only genus with flexible encasing scales 

 and a rudimentary fin-ray element (small in the single 

 species examined) in which the supraoccipital does not 

 extend far forward and is without a crest, although the 



176 



