smaller, isolated scale plates with the same kind of sur- 

 face granulations on the head, especially dorsally and 

 ventrally, and on the body posterior to the carapace in 

 the region between the soft dorsal and anal fins and on 

 the caudal peduncle. The scale plates, and apparently 

 thick skin, on the caudal peduncle completely obscure 

 the vertebral column. 



The possible phylogenetic implications of the presence 

 of ostracioidlike scales in the triacanthodid 

 Protobalistum is discussed in the section on subfamilial 

 relationships. 



Spinacanthus cuneiformis (Blainville 1818:359), a 

 single plate, MNHN (LP) 10918, 104 mm SL, presently is 

 a very incomplete and poorly preserved specimen which 

 appears to have lost some of its surface and bony struc- 

 ture since it was studied by Volta (1796), Blainville 

 (1818), and Agassiz (1839 illustration, 1844c description). 

 Much less can be seen in the specimen today than in the 

 illustrations of it by Volta and Agassiz. The spiny dorsal 

 fin is even more enormous than in Protobalistum, with 

 the first spine, even with a slight amount of the distal tip 

 missing and not measurable, being 95.7 mm (92.0' r 

 SL). The second to fourth dorsal spines are missing too 

 much of their distal ends to be measureable, but ap- 

 parently they were almost as long as the first spine. The 

 fifth spine is about 35 mm (34% SL) long, and a smaller 

 sixth spine, if present, cannot be detected. There are 

 large retrorse thomlike barbs along the anterior edge of 

 at least the basal third of the first dorsal spine. There 

 appear to have been about 10 fin rays in both the ap- 

 posed soft dorsal and anal fins. The slightly rounded 

 caudal fin is poorly preserved, but it had about 12 rays 

 (Agassiz thought 7 or 8 branched rays plus an un- 

 branched ray above and below). The pectoral fin has 

 about 12 or 13 rays (Agassiz thought 12). Agassiz es- 

 timated the number of vertebrae to be no more than 20, 



and I would guess at about 20 judging on the basis of the 

 size of those that show and of the space available for 

 those that do not show. The length of the caudal pedun- 

 cle is 26.0 mm (24.9% SL) and its depth at least 7.0 mm 

 (6.7% SL). The greatest body depth is 39.0 mm (37.4% 

 SL). The eye is placed high in the head just below the 

 base of the first dorsal spine and has a diameter of 7.2 

 mm (6.9% SL). 



The branchiostegal rays from both sides are evident, 

 and in the more clearly seen set there appear to be six 

 rays. The prefrontal is one of the few bones that are rela- 

 tively well preserved, and it has an appropriately 

 moderate size. Agassiz thought that a pelvic spine was 

 present and displaced in the region of the branchios- 

 tegals. At least at present I find no structure in the 

 branchiostegal region that could possibly be interpreted 

 as a pelvic spine. However, just posterior to the branchi- 

 ostegals and below the pectoral fin is a vague indication 

 of a long rodlike strut of bone oriented vertically. It is 

 possible that this is the structure seen by Agassiz, and it 

 might possibly be a thin, weakly developed barbless 

 pelvic spine. It would appear to me more likely to be a 

 part of the pectoral girdle, perhaps the thickened pos- 

 terior edge of the coracoid. As with Protobalistum, it is 

 not known with surety whether Spinacanthus had a 

 pelvic fin. 



The teeth in the lower jaw are well-preserved. There 

 appear to have been about seven of them in the left den- 

 tary. A few of these teeth are especially fully exposed and 

 can be seen clearly in three dimensions. One of these is 

 2.0 mm (1.9% SL) long from its flat circular base to its 

 distal tip. The tooth tapers only very slightly from basal- 

 ly to distally throughout most of its length, but in the 



Figure 16.— Spinacanthus cuneiformis: 



lateral view of holotype, 104 mm SL, 



Eocene of Monte Bolca, Italy. 



