fornasinii-like ancestor, which may have been higher 

 crested than the Recent Lactoria. 



In the Ostracion-Rhynchostracion line of ostraciins 

 derived from a probably pentangular form, there seems 

 to have been a tendency for the distance between the 

 ventrolateral ridges to be decreased to about the same as 

 that between the dorsolateral ridges, while the dorsal 

 carapace crest tended to be reduced in height, leading to 

 a basically rectangular form with overtones of the 

 ancestral pentangular form depending on the degree of 

 development, if any, of the dorsal crest. Ostracion and 

 Rhynchostracion probably evolved from the same pre- 

 Lactoria comuta-\ike line that gave rise to Lactoria and 

 Tetrosomus. The Ostracion-Rhynchostracion line is en- 

 visioned as diverging from them in the shape of the 

 carapace and in the loss of spines from the probably 

 spiny carapaced ancestor, while developing large last 

 basal pterygiophores with large carapace supporting 

 shelves in both the dorsal and anal fins and reducing the 

 size of the haemal spines of the four postanal vertebrae, 

 with the haemal spines losing the foramen and being 

 solid. 



Ostracion shows the extreme in dorsal carapace crest 

 reduction and increased distance between the dorso- 

 lateral ridges, the carapace being essentially rectangular, 

 while it retains the most generalized ostraciid condition 

 of the anterior vertebral fusion complex. In Rhyncho- 

 stracion, however, one species {nasus) retains a rela- 

 tively well-developed dorsal carapace crest while in the 

 other (rhinorhynchus) it is obsolete, the carapace being 

 almost as rectangular as in Ostracion, with both species 

 of Rhynchostracion having the anterior four vertebrae 

 fully fused to one another. Rhynchostracion has the addi- 

 tional slight specialization of a bulbous or protruding 



snout. Ostracion and Rhynchostracion can be viewed as 

 closely related sibling genera from a common ancestral 

 group, with Rhynchostracion slightly more specialized in 

 the greater anterior vertebral fusion and expanded 

 ethmoid region while retaining a slightly more general- 

 ized carapace shape. 



The Eocene Eolactoria is impossible to place with con- 

 fidence in either one or the other of the two subfamilies 

 recognized here for the Recent species, simply because its 

 dorsal, anal, and pectoral fin rays are not preserved and 

 none of the internal features equally diagnostic between 

 the two subfamilies can be seen (Tyler 1973a). The 

 presence of enormous preorbital and postanal carapace 

 spines in Eolactoria would seem to relate it to either the 

 Atlantic Acanthostracion (Lactophrysinae) or the Indo- 

 Pacific Lactoria (Ostraciinae), from both of which it 

 differs in the immensity of the elongation of the spines 

 (equal to the standard length of the fish) and in the 

 presence of a much shorter unpaired spine directed 

 anteriorly from between the front of the eyes, the latter 

 unique in the family. Without knowing the appropriate 

 critical diagnostic characteristics, Eolactoria is simply 

 placed here questionably in the Lactophrysinae, the 

 more generalized of the two subfamilies, on the theory 

 that an Eocene ostraciid is more likely to be generalized 

 in most respects than specialized. It is possible that 



FifTure i79.— Eolactoria aorbinii: lateral view 



of holotype, 15.5 mm SL, Eocene of Monte Boica, 



ItalyCIVlerigrSaifig. 1). 



242 



