48 DR. W. a. EIBEWOOD OlST THE 



been introduced into the literature of the subject by von Ihering's 

 statement (9. p. 229 & p. 233) that Scymnus departs from the 

 condition found in Acanthias and Scyllium by having no double 

 vertebrse in the tail, or only one or two vertebrae with imperforate 

 crural and intercrural plates ; for Gradow (4. p. 195) has repeated 

 the statement iu his memoir on the vertebral column of fishes, 

 "without having observed that Mayer (13. p. 265) had corrected 

 von Ihering on this point. The figure given by Mayer (13- pi. 18. 

 fig. 9) shows that normal diplospondyly commences at the forty- 

 third vertebra in Scymnus licJiia. In a specimen of this species, 

 wliich by the kindness of Mr. Gr. A. Boulenger I was allowed to 

 examine at the Natural History Museum, I found that nine 

 myomeres of the tail were diplospondylous. The first of these 

 segments ^vas l^ituated immediately behind the cloaca, and the 

 hjBmal arches commenced at the same place. The relations were 

 thus exactly those which occur in other Selachians. 



The passage from the diplospondylous condition in the tail to 

 the monospcndylous condition in the trunk is not abrupt, but 

 gradual ; and the four or five vertebrae involved in the transition 

 ofler an irregularity which deserves more notice than has hitherto 

 been accorded to it. Although three theories at least have been 

 propounded to explain the origin of the caudal diplospondyly, it 

 does not appear to have occurred to anyone to inquire minutely 

 into the detailed structure of these transitional vertebrae; yet 

 they hold, as it w^ere, the key to the situation. The transititn 

 is invariably in the proximity of the cloaca, as von Ihering and 

 Mayer (9. p. 228, and 13. p. 261) have already pointed out. It 

 occurs behind the last rib-bearing vertebra, and in front of the 

 vertebra carrying the third or fourth complete hsemal arch (see 

 fis^'. 2). As a rule it does not extend through more than four or 

 five myomeref!, but according to Mayer (13. p. 266) six body- 

 segments are involved in Mustelus. 



The transitional vertebrae do not stand in any constant relation 

 with the dorsal fins, for they are anterior to the first dorsal fin 

 in Ehina (13. pi. 18. fig. 1), below it in Scylliiim, between the two 

 dorsals in Acanthias, and below^ the second dorsal in Scymnus. 

 Since, however, the dorsal fins are variable in position with respect 

 to the cloaca in different genera, and the transitional vertebrae 

 are definitely related to the position of the cloaca, it is but a 

 logical conclusion that the vertebral transition shall not be related 

 to the position of the dorsal fins. Since caudal diplospondyly 



