CAUDAL DIPLOSPONDTLT OP SHAUSS. 57 



the metamerism of the muscular and nervous system is perfect. 

 As a rule the segmentation of this sheath is determined by the 

 myomeres, iu the manner explained by Balfour (1. p. 453), so 

 that the vertebra? are as numerous as the myomeres; but there 

 is nothing to prevent the vertebral segments being twice, or even 

 three times, as many as the primary segments. To have frac- 

 tional parts, however, is obviously impossible. Even allowing 

 that the transition between the trunk and tail regions is beauti- 

 fully gradual, yet, as will be seen by a glance at figure 2, the 

 last monospondylous skeletal segment (3) is followed imme- 

 diately by one with two centra and tw^o crural plates (4 " 5). 



The most logical conclusions, therefore, from the facts at 

 command, are, that the condition of diplospondyly in the tail of 

 Sharks is secondary, but of ancient date ; and, further, that it is 

 purely adaptive, being calculated to maintain a due proportion 

 between length of centrum and width of body, without dimin- 

 ishing the length of the muscle-segments. In the region of the 

 body from the cloaca to the caudal fin, the demand for increased 

 flexibility is prepotent over the normal tendency of the chon- 

 drified chordal sheath to segment in such a way that the centra 

 are as numerous as the myotomes. 



This, of course, is not an explanation of diplospondyly, it is 

 merely a suggestion for its raison d'Ure. That the diminution 

 in the length of the tail which would be entailed by a shortening 

 of the myomeres, and consequent restoration by this means of 

 the balance between the length of the centra and the width of 

 the body, would be a disadvantage, is also a pure assumption, 

 Indeed, the study of Teleostean fishes shows that a shortening of 

 the tail by the abolition of the terminal vertebrae may, and does 

 occur; and this without any compensating increase in the 

 skeletal parts, for the caudal segments of Teleosteans are mono- 

 spondylous. But, in this connection, Amia comes to our assist- 

 ance; for here, in spite of a homocercal tail and presumably 

 abbreviated vertebral column, the segments of the caudal axis 

 occur two to each myomere and neuromere (9. p. 231). And, as 

 in Selachians, the last segments of the body, namely those in the 

 hinder part of the caudal fin, are monospondylous (Pranque, 3 ; 

 Kolliker, 12; Shufeldt, 16; Hay, 8 ; and others). 



"Whether, therefore, we accept the view of Pranque (3. p. 10) 

 that in Amia those vertebral bodies of the tail which are destitute 

 of neural and haemal arches have been intercalated between the 



LINJSr. JOURN. — ZOOLOGY, VOL. XXVII. 5 



