24 OSTEOLOGY OF SCOMBROID FISHES 



spines are attached at about the anterior third of the length of the 

 pelvic bones.* 



The branchial and hyoid elements are almost identical with those 

 of Promethichthys. 



There are 41 abdominal vertebra?, and in the specimen at hand 62 

 caudal, besides a few (probably not exceeding 5) that are missing at 

 the end of the tail. 



Lepidopus differs from Promethichthys in having no abdominal 

 haemal arches. The last two abdominal vertebrae bear small, slender 

 parapophyses, but the ribs are still attached to the vertebral centra at 

 the base of them, though they are also attached to the full length of 

 the parapophyses at the side. The abdominal haemal arches and spines 

 in Promethichthys are not differentiated from those of the caudal re- 

 gion, but in Lepidopus the change from the small parapophyses of the 

 last abdominal vertebrae to the large haemal arch and spine of the first 

 caudal vertebra is very abrupt. The ribs are set in very deep pits low 

 on the vertebral centra, and, as in Promethichthys, each rib or haemal 

 spine occupies a positon at about the middle of the length of its verte- 

 bra. There is a sharp, low, lateral ridge running the length of the 

 vertebral column. Along the caudal region it occupies the middle of 

 the sides of the vertebra?, and is a little elevated at each end of each 

 vertebra. On the abdominal region it is ventral in position, and forms 

 a sharp lower edge, which bears the sockets for the ribs. 



The first two vertebra? bear epipleurals only. There are only three 

 other epipleurals attached to the vertebral centra just above the base 

 of the ribs. There are no long intermuscular bones incasing the body, 

 as in Promethichthys. The hypural is small, but in no way remark- 

 able; the caudal rays are not deeply divided. The interneurals are 

 1-shaped, with the horizontal limb connecting them, and forming the 

 outline of the back. They are only slightly expanded laterally just be- 

 neath the skin. No ossified, separate baseosts are present. The fourth 

 and fifth interneurals bear large bean-shaped swellings near the upper 

 end, each nearly as large as a vertebra. 



*The abdominal ventrals might indicate a direct descent from more primitive 

 forms with similar ventrals, did not Promethichthys (with thoracic ventrals) so 

 obviously stand between. This condition in Lepidopus probably shows an independ- 

 ent degeneration or modification. As the body has become elongate, and the ven- 

 trals functionless, the latter have lost their connection with the shoulder girdle and 

 have moved backwards. Since the above was written a paper by Mr. C. Tate Regan 

 (Ann. and Mag. Nat. Hist. (8) Vol. III. 1909) has been received in which he has 

 come to the same conclusion : "This condition may be regarded as secondary and 

 due to the degeneration of the fins." 



