4.5% SL) is slightly greater than that between the last 

 few spines (4.1 mm, 3.6% SL). The anal fin has 14 rays, 

 the distal ends of those that show being branched. The 

 elongate rounded caudal fin has 12 rays, the uppermost 

 ray and the lowermost ray unbranched and the inter- 

 vening 10 rays branched. The pectoral fin rays cannot be 

 counted, nor are any pelvic fin rays evident, although a 

 small ray or two could well be present but obscured by 

 the pelvic spine. The branchiostegal rays are 2 + 4 = 6, 

 and the first two rays are not much wider than the others. 

 There are clearly 12 caudal vertebrae and probably 8 ab- 

 dominal vertebrae. The supracleithrum is placed ver- 

 tically in relation to the axis of the skull. The 

 parasphenoid is relatively straight in the region of the or- 

 bit, and the ventral flange seems to have been only 

 moderately developed, probably only slightly deeper 

 than the depth of the shaftlike portion of the bone above 

 it. 



The prefrontal is very large and sturdy, and has at 

 least a moderate anteroventral extension alongside the 

 lower region of the ethmoid, although it is not clear 

 whether this extension took the form of a long process 

 which also sutured anteriorly with the posterolateral 

 wing of the vomer, as in Recent triacanthids. The upper 

 half of the prefrontal apparently was well separated from 

 the upper half of the ethmoid by a block of cartilage 

 larger than in the Recent species. The ethmoid appears 

 to have been large and sturdy, with a steeply oblique 

 anterior face that then tapered posterodorsally to suture 

 with the frontal and prefrontal, much as illustrated by 

 Tyler (1968:361, fig. 206) for Triacanthus biaculeatus, 

 but with the frontal extending a little further forward. 



The upper jaw apparently is unnaturally displaced 

 posteriorly, for the tip of the upper jaw is well behind 

 that of the lower jaw, a functionally unlikely arrange- 

 ment in this family. As presently placed, the posterior 

 end of the premaxillary pedicel abuts against the top of 

 the oblique anterior face of the ethmoid just anterior to 

 the anterior end of the frontal. In life, however, the 

 pedicel undoubtedly never was retracted this far postero- 

 dorsally and probably slid along the anterior basal region 

 of the ethmoid when the mouth was protracted, as in the 

 Recent species. The moderate-sized eye is located just 

 above the middle of the distance between the snout and 

 the spiny dorsal origin. The pelvis posterior to the pelvic 

 spines appears to be a stout shaft of bone. The anterior 

 and lateral surfaces of the first dorsal spine and the 

 anterior, dorsal, and ventral surfaces of the pelvic spine 

 are covered with low spiny processes, probably asperities 

 on the scales covering these spines. 



The caudal fin supporting structures are not well 

 preserved, but grooves indicate that there was probably a 

 good separation between what in generalized tria- 

 canthoids are the second and third hypurals. The condi- 

 tion of the parhypural, epurals, and uroneural elements, 

 if present and separate, is unclear. The first basal 

 pterygiophore of the spiny dorsal fin (supporting the first 

 two spines) has a strong ventral shaft directed ventrally 

 to between the middle to basal region of the skull and 

 what is probably the neural spine of the first vertebra 

 closely applied to the skull. The second basal pterygio- 





HI 



-\.)r 



Figure 43.— Protocant/iodes ombonii: lateral 

 view of abdominal region (head to left) to 

 show the pelvis and pelvic spine (one from 

 each side), holotype, 112 mm SL, Eocene of 

 Monte Bolea. Italy. 



phore (supporting the third spine) has its shaft angled 

 slightly anteroventrally and in contact with the distal 

 end of the neural spine of the second vertebra. The third 

 and fourth basal pterygiophores have their shafts 

 directed ventrally, about vertical to the vertebral 

 column, and in contact, respectively, with the distal ends 

 of the neural spines of the third and fourth vertebrae. If 

 a fifth basal pterygiophore supporting the slender, 

 minute sixth spine was present, it apparently was very 

 small, for it does not show in either plate. The first basal 

 pterygiophore of the soft dorsal fin is placed between the 

 neural spines of the sixth and seventh abdominal 

 vertebrae, and there appears to have been an approxi- 

 mately one to one ratio between the number of dorsal and 

 anal fin rays and their basal pterygiophore supports. 



The teeth are relatively large incisors, about five in 

 each half of each jaw judging from the size of the 

 somewhat scattered elements. It is not possible to tell 

 whether an inner series of teeth was present in addition 

 to the major outer series. The largest teeth, antero- 

 medially in each jaw, have stout bases toward the sockets 

 but taper into blunt nipplelike caps distally, while the 

 teeth more laterally in each jaw have lesser cusps and, in 

 some cases, almost rounded distal edges. 



Protacanthodes differs from the other triacanthids 

 mainly by having: 1) a deep and only slightly tapered 

 caudal peduncle; 2) an extremely elongate rounded 

 caudal fin; 3) the spiny dorsal fin base slightly longer 

 than the soft dorsal fin base; 4) better developed spiny 

 dorsal fin basal pterygiophores whose ventral shafts are 

 slightly differently arranged; 5) the upper shaftlike por- 

 tion of the parasphenoid in the region of the orbit rela- 

 tively straight and its ventral flange only slightly deeper 

 than the depth of the shaftlike portion; 6) the lower two 

 branchiostegal rays not enlarged; 7) possibly less fusion 

 of the hypurals; 8) one more abdominal vertebra anterior 

 to the first basal pterygiophore of the soft dorsal fin; 

 9) and in having scales with a row of upright spinules. All 

 of these features of difference between Protacanthodes 

 and Recent triacanthids relate Protacanthodes to the 



