1'OTTIIOFF an.] KKI.I.KV: OSTKOLOOICAI, I)K VKI.OI'M KNT IN SWOKDFISH 



fins (Leim and Scott 1966; Ovchinnikov 1970), 

 but larvae and juveniles have one continuous 

 dorsal and anal fin (Nakamuraetal. 1951; Yabe 

 et al. 1959). During development the fin rays in 

 the center of the fins stop growing and the rays 

 become subcutaneous. The subcutaneous rays 

 and their pterygiophores are present in the 

 adults and were dissected in our largest 668 mm 

 SL specimen. In three scombrid genera, 

 Scomber, RastreUiger, and Auxis, we find a first 

 dorsal and second dorsal fin separation similar to 

 that in adult Xiphias, except that in these 

 scombrids the two fins are separate initially even 

 though the first and second dorsal fin pterygio- 

 phores are continuous (Kramer 1960; Potthoff 

 pers. obs. on Auxis). There is only one anal fin in 

 these three scombrid genera, whereas adult 

 Xiphias have two anal fins. 



All dorsal rays in Xiphias are bifurcated at 

 their bases (Figs. 14, 15) as in Coryphaena 

 (Potthoff 1980). This probably is not the case in 

 most perciforms where the spinous rays of the 

 first dorsal fin have a closed base with a foramen 

 and the distal radials are situated outside the 

 bases of the first dorsal fin spinous rays (Kramer 

 1960; Potthoff 1974, 1975; Potthoff et al. 1980). 



The anteriormost dorsal pterygiophores in 

 Xiphias insert in the second interneural space 

 (Figs. 11, 13), asinthegempylidsandtrichiurids 

 (Potthoff et al. 1980), but not as in the serranids, 

 sparids, apogonids, scombrolabracids, and 

 scrombrids where the anteriormost pterygio- 

 phores insert in the third interneural space 

 (Matsui 1967; Fraser 1972; Potthoff 1974, 1975; 

 Houde and Potthoff 1976; Fritzsche and Johnson 

 1980; Potthoff et al. 1980), and not as in the 

 coryphaenids in which they insert in the first 

 space (Potthoff 1980). No predorsal bones were 

 present in Xiphias. All scombrids and most 

 scombroids also lack predorsal bones, however 

 some gempylids, e.g., Ruvettus (Potthoff et al. 

 1980), have one predorsal. Most other perci- 

 formes have predorsals in the first and second 

 interneural spaces. 



The first dorsal pterygiophore in Xiphias is 

 variable in development (Figs. 7, 8) and 

 originates either from one or two pieces of 

 cartilage. In scombrids (Potthoff 1974, 1975), a 

 two-part development of the first dorsal pterygio- 

 phore was not evidenced, but in Morone it was 

 (Fritzsche and Johnson 1980). 



The last (posteriormost) pterygiophore of 

 Xiphias has a serially associated double ray and 

 a stay (Figs. 9, 10). In Xiphias, as probably in all 



Perciformes, the stay develops from the prox- 

 imal radial cartilage. The stay is not posteriorly 

 bifurcated as in most scombrids (Potthoff pers. 

 obs.), nor does it ossify into two parts as in most 

 gempylids and some trichiurids (Potthoff et al. 

 1980). 



Xiphias lacks middle radials as does Cory- 

 phaena (Potthoff 1980), whereas many Per- 

 ciformes probably have middle radials at least 

 for some of the posteriormost dorsal and anal 

 pterygiophores (Kramer 1960; Berry 1969; 

 Potthoff 1974, 1975; Houde and Potthoff 1976; 

 Potthoff et al. 1980; Fritzsche and Johnson 1980). 



In Xiphias the caudal rays are supported by 

 only two centra (urostyle and preural centrum 2) 

 (Figs. 17, 21, 22). This is unusual, because in most 

 perciforms three centra support the caudal rays 

 (Berry 1969; Houde and Potthoff 1976; Potthoff 

 1980; Potthoff et al. 1980; Fritzsche and Johnson 

 1980), and in most scombrids four or five centra 

 support the caudal rays (Collette and Chao 1975; 

 Potthoff 1975; Collette and Russo 1978), except in 

 Scomber and RastreUiger where three centra 

 support caudal rays (Potthoff pers. obs.). 



Xiphias lacks a second uroneural in the caudal 

 complex which is present in the basic perciform 

 caudal such as in Archosargus (Houde and 

 Potthoff 1976), Elagatis (Berry 1969), Scom- 

 brolabrax (Potthoff et al. 1980), Morone 

 (Fritzsche and Johnson 1980), and Coryphaena 

 (Potthoff 1980), but is absent in the scombrids 

 (Potthoff 1975). The single uroneural of Xiphias 

 does not fuse to the urostyle in adults as in 

 Thunnini and Sardini (Collette and Chao 1975; 

 Potthoff 1975; Collette and Russo 1978), but in 

 several specimens anomalous shapes of the 

 uroneural were observed (Fig. 22). 



We believe that Xiph ias has lost preural cen- 

 trum 3, because a centrum having an autogenous 

 haemal spine and a neural spine with articular 

 cartilage is lacking (Figs. 20-23). However, 1 

 specimen out of 164 examined with the unusual 

 vertebral count of 16+11=27 (typical counts 15+ 

 11 or 16+10=26) had two autogenous haemal 

 spines on preural centra 2 and 3. To our knowl- 

 edge, a perciform caudal with only one autogen- 

 ous haemal spine as in Xiphias has not been 

 reported previously. We cannot totally rely on 

 Monod (1968) or any other osteological descrip- 

 tive work dealing only with adult fish because 

 Potthoff (1975) showed that some autogenous 

 hypural parts fuse during development and can- 

 not be recognized in adults. 



There is considerable fusion of caudal complex 



183 



