Contrary to Tyler (1970b:29), at least the more exten- 

 sively branched dorsal, anal, and pectoral fin rays of all 

 molids, and in Ranzania the pseudocaudal rays as well, 

 often bear cross-striations distally, the number of cross- 

 striations probably increasing with increasing specimen 

 size. The molid cross-striations, however, are fewer in 

 number and more distally placed than in the fins of other 

 plectognaths. 



In Mola the pseudocaudal fin is supported by the last 

 three caudal vertebrae, of which the last is a simple rod 

 representing only a centrum. The penultimate vertebra 

 has short neural and haemal spines which curve ante- 

 riorly to be supported at their distal ends by the fully 

 developed neural and haemal spines of the antipenul- 

 timate vertebra, while their posterior edges support the 

 more medially placed radial elements. But it is the 

 neural and haemal spines of the antipenultimate 

 vertebra which support most of the radial elements. 

 There are seven radial elements above and eight below, 

 and 15 pseudocaudal fin rays, in both specimens examin- 

 ed. Fraser-Brunner (1951) gave 16 rays as the norm for 

 the south Pacific M. ramsayi and 12 for the worldwide M. 

 mola, with which distinction the two California 

 specimens of M. mola examined here are not in agree- 

 ment. 



Masturus has one less caudal vertebra than Mola, and 

 it is obvious that Masturus has lost what is the last 

 vertebra oi Mola, the simple rod representing a centrum. 

 With the exception of Masturus lacking this rodlike cen- 

 trum the pseudocaudal supports of Mola and Masturus 

 are rather similar, the ultimate vertebra oi Masturus be- 

 ing homologous and functionally corresponding to the 

 penultimate of Mola, and the penultimate of Masturus 

 to the antipenultimate of Mola. 



The most diagnostically important difference between 

 Masturus and Mola in this region is that Masturus has 

 more (20 or more) rays in the pseudocaudal fin than does 

 Mola, the additional rays mostly being accounted for by 

 those in the medial lobe that is present in Masturus but 

 absent in Mola, and that bony ossicles do not develop 

 distally on any of the pseudocaudal rays in Masturus but 

 do in Mola. Whereas at least most of the rays in the up- 

 per and lower halves of the pseudocaudal fin of both 

 Masturus (exclusive of the medial lobe) and Mola are in- 

 dividually supported by radial elements, at least most of 

 the rays of the medial lobe of Masturus are not supported 

 by radial elements but rather take support from the last 

 vertebral centrum. In the single specimen of Masturus 

 examined there are about (it being difficult to decide 

 where the pseudocaudal rays begin) 7 rays in the upper 

 half of the pseudocaudal, 10 below, and 6 in the medial 

 lobe, with the dorsal fin having 20 rays and the anal fin 

 19, a total dorsal, pseudocaudal, and anal count of 62. 

 Fraser-Brunner (1951) gave a total count of 60 to 62, with 

 eight (rarely seven or nine) rays in the medial lobe for M. 

 lanceolatus, and a total of 55 to 57, with four (rarely three 

 or five) in the medial lobe for M. oxyuropterus. Obvious- 

 ly, more extensive data is needed for these various fin 

 counts in both Mola and Masturus, and such can be easi- 

 ly ascertained only from cleared and stained specimens, 

 or from very finely radiographed specimens. 



In short, Mola and Masturus differ primarily mainly in 

 the details of the pseudocaudal fin and its sup- 

 ports: 1) Mola has a ninth caudal vertebral centrum 

 that is lost by Masturus; 2) Masturus has a group of rays 

 in a medial lobe of the pseudocaudal fin that are sup- 

 ported by the last vertebral centrum rather than by 

 radials, this medial lobe not being present in Mola and 

 perhaps being the remnants of true caudal fin rays; 



3) Mola develops bony ossicles at the distal end of many 

 of the pseudocaudal rays while Masturus never does; and 



4) Mola has a better developed neural spine on the eighth 

 caudal vertebra than does Masturus. Other differences 

 are: 1) Mola has a supraneural that appears in 

 Masturus to be fully fused to the first basal pterygio- 

 phore of the dorsal fin; 2) Mola has the interoperculum 

 ossified, while it is unossified in the single small 

 specimen of Masturus examined; 3) the trituration teeth 

 in the lower jaw of Mola are divided into right and left 

 series, while Masturus has a major medial series. 



The above differences between Mola and Masturus 

 seem to justify the retention of them as distinct genera, 

 even if they should be found to be represented by a single 

 species each. The far more numerous differences listed 

 between Ranzania and Mola-Masturus reflect in part the 

 ease of distinguishing between subgroups composed of 

 only a few species. In the present case each of the three 

 genera either has only a single species or only a single 

 species was examined for this work, so there is, in es- 

 sence, no intrageneric variation and the comparisons 

 made are only between three units. Thus, it seems to me 

 that the extensive differences listed between Ranzania 

 and Mola-Masturus do not necessarily justify the recog- 

 nition of two subfamilies within the family Molidae, and, 

 as a personal opinion, I do not think the differences 

 between the two groups are of sufficient magnitude for 

 subfamilial recognition. 



Generic relationships. — With it being clear that 

 Mola and Masturus are far more closely related to one 

 another than to Ranzania, the most pertinent remaining 

 question is whether one or the other of these two lines of 

 molid evolution is more generalized than the other and 

 anatomically closer to the triodontidlike ancestral group. 

 The only other question of consequence is whether Mola 

 or Masturus is the more generalized within its line of 

 radiation. 



To address the last question first. Mola is more 

 generalized than Masturus by: 1) the retention of the 

 ninth caudal vertebra, separate supremeural, and ossified 

 interoperculum; 2) by the better development of the 

 neural spine of the eighth caudal vertebra; 3) by the 

 retention of the trituration teeth in right and left series to 

 either side of the midline in the lower jaw. Conversely, 

 Masturus is more generalized than Mola by: 1) the 

 possible retention of true caudal fin rays in the medial 

 lobe of the pseudocaudal fin; and 2) the lack of ossicles at 

 the ends of any of the pseudocaudal fin rays. Overall, it 

 would seem that Mola has remained slightly more gener- 

 alized than Masturus. 



Many of the differences between Ranzania and Mola- 

 Masturus indicate that Ranzania is more specialized 



