groove in which it is received when depressed (the latter 

 character, however, being lacking in the subgenus 

 Acreichthys noted [described as new| below). On the one 

 hand it leads to the rest oi' the genera retaining the pel- 

 vic spine but having the dorsal spine behind the eye and 

 without groove; of these, Stephanolepis seems to be the 

 group round which the specialized forms such as Chaeto- 

 derma and Monacanthus are clustered. On the other 

 hand extends the large series of genera in which the pel- 

 vic spine has been lost. ..." 



I do not find that the dorsal spines and encasing scales 

 in Pervagor are better developed than in most other 

 genera, and certainly they are no better developed than 

 in Stephanolepis, Monacanthus, and Paramona- 

 canthus. Moreover, I do not think that the somewhat for- 

 ward position of the dorsal spine (over the middle of the 

 eye) can be considered the generalized condition when 

 the ancestral balistids have the dorsal spines behind the 

 eye. 



As discussed previously, the least forward migration of 

 the dorsal spines is the most generalized monacanthid 

 condition, with the spines over the rear of the eye. Per- 

 vagor does not possess the generalized monacanthid con- 

 dition of the supraoccipital, for it lacks the forward ex- 

 tension with vertical crest to the front of the eye that is a 

 remnant of the condition of the balistid supraoccipital. 



In fact, the only definitely generalized feature of Per- 

 vagor to my way of thinking is its pelvic apparatus, with 

 it having a moderate dorsal lobe, three segments in the 

 flexible encasing scale series and moderately well- 

 developed rudiments of the fin-ray element. As dis- 

 cussed previously, a groove behind the dorsal spines 

 would seem a priori a generalized condition in mona- 

 canthids since the ancestral balistids have such a groove. 

 But of the genera of monacanthids with a groove behind 

 the dorsal spines, only Pervagor has a generalized pelvic 

 apparatus, all the other genera having the segments of 

 encasing scales reduced in number and inflexible or ab- 

 sent entirely, and without a fin-ray element. 



All of the genera with the groove behind the dorsal 

 spines have the most specialized condition of the 

 supraoccipital, without a long or even moderate forward 

 extension and no crest. It seems as likely that the 

 monacanthid postdorsal groove is a de novo develop- 

 ment in moderately specialized forms as it is that it is a 

 hold over from the ancestry of the monacanthids, the 

 most generalized Recent species of which would then 

 have to be presumed to have subsequently lost it. 



A similar situation occurs with the interpretation of 

 the 2-1-4 branchiostegal number found only in a few 

 genera of monacanthids, these all being genera with a 

 specialized pelvic apparatus in which the fin-ray element 

 is absent and the segments of encasing scales are reduced 

 in number and inflexible or absent entirely, a small dor- 

 sal lobe is present on the pelvis, and there is either no 

 forward supraoccipital extension (Nauodon, Acantha- 

 luteres, Cantherhines, Amanses) or only a small exten- 

 sion to above the middle of the eye {Alutera). Thus, on 

 all of the most important grounds, the genera with the 2 

 + 4 number of branchiostegals are much more specializ- 



ed in general than are many of the genera with the 1 -t- 4 

 number, including Stephanolepis, Monacanthus, and 

 Paramonacanthus. 



Pervagor also possesses one unique feature found in 

 neither the other monacanthids nor in balistids. The 

 lateral surface of the anterior region of the pelvis has a 

 protuberance for articulation with a similar 

 protuberance from the posterior edge of the coracoid, the 

 pelvis rotating around this joint when the dewlap is ex- 

 panded. 



In short, Pervagor cannot be considered one of the 

 most generalized monacanthids, but, rather, it is 

 probably a derivative of a Paramonacanthus-like stock. 

 In both Paramonacanthus and Pervagor the dewlap is 

 short, not extending out beyond the beginning of the en- 

 casing scale series, which protrude conspicuously from 

 the ventral contour. The specializations of Pervagor 

 beyond the Paramonacanthus level are the loss of the for- 

 ward extension of the supraoccipital, the more anterior 

 position of the dorsal spines, the greater elongation of the 

 body, and the development of the rotation processes on 

 the coracoid and pelvis. 



Two other genera with a flexible encasing scale series 

 and rudiments of the fin-ray element are Laputa and 

 Chaetoderma, both of which also have the dorsal spines 

 placed in the generalized location over the rear of the eye. 



Chaetoderma has a relatively well-developed forward 

 extension of the supraoccipital and at least a low weak 

 crest, the forward extension being not much less exten- 

 sive than in Stephanolepis, Monacanthus, and Para- 

 monacanthus. The most outstanding specializations of 

 Chaetoderma are that the encasing scale segments have 

 been reduced to two, while retaining their flexibility, and 

 that the two rudiments of the fin-ray element are 

 minute. The ventral flap of Chaetoderma is small and 

 similar to that of Paramonacanthus. Chaetoderma could 

 well be a derivative of Paramonacanthus, specialized by 

 the increased depth of the body, the great development 

 of dermal flaps on the skin, the reduction in the number 

 of segments of encasing scales and in the size of the rudi- 

 ments of the fin-ray element, and the slight reduction in 

 the size of the forward extension of the supraoccipital. 



However, the scales of Chaetoderma are strongly 

 reminiscent of those of the rough skinned Monacanthus, 

 especially those of the Indo-Pacific species, in which the 

 strong central upright spinule is backwardly directed and 

 bears a notch along its anterior or outer edge. The scales 

 of Chaetoderma are an elaborate version of this, with a 

 deep cleft in the usually strong single spinule dividing 

 the spinule into a smaller upwardly or anteriorly directed 

 prong and a larger posteriorly directed prong. 



I suspect that the Australasian Chaetoderma is deriv- 

 ed from the Indo-Pacific stock of Monacanthus, by all 

 the ways of specialization mentioned above in com- 

 parison to Paramonacanthus, and additionally by the 

 loss of caudal peduncle armature and the great reduc- 

 tion in the size of the dewlap probably associated in some 

 way with the camouflage value of the extensively 

 developed dermal flaps. Additional evidence for a close 

 Chaetoderma and Indo-Pacific Monacanthus relation- 



