edge of the fan in Psilocephalus are not much enlarged or 

 differentiated from those just anterior to them. The dor- 

 sal lobe is poorly developed in Oxymonacanthus, which 

 flares only a small or moderate sized fan and which has 

 no modified scales in the fan. 



Of the genera with fixed encasing scales, the pelvis 

 ranges from very strongly (e.g., Amanses) to only moder- 

 ately (e.g., Oxymonacanthus) developed. Of the genera 

 with no encasing scales, none have a dorsal lobe, and the 

 size of the pelvis ranges from massive (e.g., Alutera), to 

 normal (Brachaluteres), to slightly reduced in size 

 {Pseudaluteres and Acanthaluteres), to short and 

 somewhat slender (Paraluteres) , and to long but very 

 weak and slender {Psilocephalus}. 



In Pervagor alone the lateral surface of the anterior 

 region of the pelvis has a lateral knoblike expansion 

 which articulates with a similar expansion on the pos- 

 terior edge of the coracoid, forming the point of pivot 

 around which the pelvis is rotated downward and 

 forward to flare the fan. 



There is little variation in the branchial apparatus in 

 the monacanthids, except for the number of branchios- 

 tegal rays. All species with the exception of Psiloceph- 

 alus barbatus have three basibranchials, three hypo- 

 branchials, five ceratobranchials (the last toothless), 

 four epibranchials, and two pharyngobranchials (both 

 toothed). In Psilocephalus there are only two basi- 

 branchials and three epibranchials, and all of the various 

 pieces in the arches are more or less reduced in 

 massiveness, although the ceratobranchials are more 

 elongate. 



There are six branchiostegals in a 2 -(- 4 arrangement in 

 the various species examined of Nauodon, Acantha- 

 luteres, Alutera, Cantherhines, and Amanses, but only 

 four branchiostegals, in a 1 -(- 3 arrangement, in Psilo- 

 cephalus. The great majority of species examined, 

 however, have five branchiostegals, in a 1 -)- 4 arrange- 

 ment, these being of the genera Monacanthus, Stephano- 

 lepis, Paramonacanthus, Pervagor, Laputa, Rudarius, 

 Oxymonacanthus, Chaetoderma, Paraluteres, Bracha- 

 luteres, and Pseudaluteres. When five rays are present, it 

 is the second of the anterior group of two that has been 

 lost, while in Psilocephalus, with four rays, one has also 

 been lost from the posterior series of four. It is of interest 

 that of 12 specimens of Monacanthus ciliatus examined, 

 11 had the 1-1-4 arrangement on both sides, but one had 

 1 -(- 4 on one side and 2 -f 4 on the other, the second ray of 

 the anterior group being present on one side only. 



There are great differences in the number, size, and 

 placement of epipleural bones in monacanthids, which 

 has led to confusion. Fraser-Brunner (1941b:176), in his 

 diagnoses of the Balistidae and Monacanthidae, stated 

 that in balistids epipleurals were found on no more than 

 the first two caudal vertebrae, while monacanthids had 

 epipleurals on the first four or five caudal vertebrae. 

 From the materials examined here, it appears that most 

 balistids have epipleurals on the second abdominal to 

 the first caudal vertebra, but in the species of the genus 

 Rhinecanthus they extend back from the second ab- 

 dominal to the fifth or even sixth caudal vertebra. In 



most monacanthids the epipleurals begin on the second 

 abdominal vertebra and usually but not always extend 

 onto the caudal vertebrae. However, in a number of 

 species the epipleurals are confined to the abdominal 

 vertebrae. In the species of Alutera, for example, the 

 epipleurals extend from the second abdominal only to 

 the sixth (next to last) or, rarely, seventh (last) ab- 

 dominal vertebra, while in Pseudaluteres they are pres- 

 ent from the second to the eighth (next to last) ab- 

 dominal vertebra and in Psilocephalus from the second 

 to sixth (next to last) abdominal vertebra. As an ex- 

 amination of the lateral view illustrations of monacan- 

 thids presented here will show, the epipleurals of other 

 genera extend posteriorly variously anywhere from the 

 first to fifth caudal vertebrae, while in Rudarius (both er- 

 codes and minutus) and Brachaluteres they are present 

 only from the third abdominal to the second to fourth 

 caudal vertebrae, being absent on the second abdominal 

 vertebra. Thus, here is no absolute familial distinction 

 concerning the epipleurals between the balistids and 

 monacanthids. 



The epipleurals in monacanthids are usually slender 

 splints, less sturdy than in most balistids, but they are 

 ver\' sturdy in at least the single large adult oi Amanses 

 examined, and relatively well developed with large sur- 

 face area in Brachaluteres and Alutera. In Alutera the 

 epipleurals become hyperostotic in large adults. 



True ribs have previously been found among plec- 

 tognaths only in the primitive gymnodont Triodon 

 macropterus (see Tyler 1962a:798 for discussion), 

 and in the monacanthid scleroderm Pseudaluteres nasi- 

 cornis as well (see Tyler 1973b). The latter would be 

 a less amazing occurrence of ribs, in light of the ances- 

 tral balistids lacking true ribs, were Pseudaluteres at 

 least a primitive monacanthid rather than being, as 

 discussed under generic relationships, surely one of the 

 most specialized. 



The genetic information on constructing ribs would 

 appear to have been retained in the otherwise ribless 

 monacanthids and their ancestral balistids and their 

 ancestral and equally ribless triacanthids and triacan- 

 thodids from the doubtless rib bearing basal group of 

 plectognaths from which both the scleroderms and gym- 

 nodonts evolved. What peculiarity in the functional life 

 of Pseudaluteres has made it advantageous to have ribs 

 when other monacanthids (and Recent scleroderms) do 

 not is a mystery to me, but it would not appear to be as- 

 sociated with its two other primary specializations; the 

 far forward placement of the dorsal fin spine in front of 

 the eye or of the lack of encasing scales and loss of the 

 rudimentary pelvic fin-ray element. 



A posttemporal is present in all monacanthids examin- 

 ed except Psilocephalus barbatus, none of the three adult 

 specimens of which shows any evidence of one. Whether 

 the ossification center for the posttemporal is lost or in- 

 corporated indistinguishably with that of the pterotic 

 remains problematical. The dorsal head of the supra- 

 cleithrum in Psilocephalus articulates directly with the 

 lateral surface of the pterotic. This loss of the posttem- 

 poral is undoubtedly correlated with the greatly reduced 



167 



