1032 REPORT OP COMMISSIONER OF FISH AND FISHERIES. [52] 



The prolongation of tlie central rays of the caudal of the young- of Mola 

 looks at first as if it represented the lophocercal prolongation of the so- 

 matic axis of the larva, but a careful examination shows that there is no 

 prolongation of the chorda, nor is the skeleton extended in any way into 

 this median projection of the caudal. If such is the case, sections alone 

 could be depended upon to clear up the question, though I am very de- 

 cidedly of the opinion that even sections will not show any evidence of 

 rudiments of the chorda in this median posterior prolongation of the 

 caudal fin. If any rudiments of a prolonged somatic axis are present, 

 it is most i)robable that the chorda will be found present. This is a 

 point which my researches cannot determine with certainty, since it is 

 desired to preserve the materials now in my possession without mu- 

 tilation for reference. Very careful examinations have convinced me, 

 aided as 1 have been with an excellent triplet, that the embryonic axis 

 oriihordiiis not prolonged into the central prolongation of the tail, 

 already described, because this median extension is a secondary forma- 

 tion, and not a part of the primitive larval axis. 



The morphology of the tail of the Molina cannot, however, be under- 

 stood without making certain assumptions, which are unequivocally 

 justified by the structure of the tail of the adult. We find, in fact, that 

 the tail of the il/o/a, when full grown, is constituted of rays, which are 

 supported entirely by interneural and interhtemal pieces, and not di- 

 rectly by any haemal and neural spines. These, however, are indirectly 

 brought to serve as supports for the interneural and interhaemal ele- 

 ments in a way which is without a parallel amongst Teleost fishes. The 

 sixteenth vertebra has its neural and haemal spines greatly prolonged, 

 and the interneural. and interhjieinal elements of the caudal fin of Mola 

 are proximally in contact with the spines of the penultimate vertebral 

 segment. 



The larva of Mola is not known, and it is therefore premature to do 

 more than surmise, on the basis of the principles of comijarative mor- 

 phology, what that larva must be like. We find that the vertebral col- 

 umn of Mola ends abruptly with a "floating" vertebra, which is cylin- 

 drical in form and evidently does not represent the last bony segment 

 as in normal Teleosts. The vertebral segments of Mola do not gradu- 

 ally become less in diameter in so striking a manner as in other Tele- 

 osts, so that the column does not terminate in an acuminate upturned 

 urostyle, or in an attenuated extension of the chorda, as in normal 

 forms. In Masturus it is not improbable that evidence of the greater 

 posterior extension of the vertebral column will be found, as would be 

 indicated by its longer caudal fin. It is evident, nevertheless, that a 

 part of the larval axis of Mola has been suppressed, for the reason that 

 the last vertebral centrum is but slightly smaller than the ijenultimate 

 one, and has evidently been formed in the skeletogenous investment of 

 the chordal axis of the larva far in advance of its posterior extremity 

 in the youngest larval stage of development. The presumption, there- 



