24 PROF. T. JEFFEET PAEKEB OX THE 



ligament ; in others two occur in a single interneural space, both being unconnected 

 with a neural spine. In the posterior part of the body there is an average of three 

 interspinous bones to each vertebra, and here, again, one of these is sometimes directly- 

 connected with a neural spine, while in other cases all three lie in an interneural space. 



The last interspinous bone (fig. 25, i. s]). 205), having to take part in the support of 

 but one fin-ray, has only its anterior arms developed, the posterior arm being represented 

 by a mere tubercle; it is situated over the ninety-second vertebra, about halfway 

 between its middle and its hinder end. 



The neural spines and interspinous bones are united by a strong longitudinal ligament 

 (figs. 20 and 25, Ig), formed by the intersection of intermuscular septa, which extends 

 from the tip of the first to that of the last neural spine ; a similar ligament (fig. 25, Ig^) 

 passes from the rudimentary posterior arm of the last interspinous bone to the upper 

 surface of the demivertebra, at about the junction of its middle and posterior thirds. 



The arrangements described in the preceding paragraphs do not appear to me to 

 give the impression that the tail has suffered mutilation ; one would hardly expect, if 

 this had taken place, to find the line of fracture through the interspinous bones so 

 much in front of that through the vertebral bodies. Moreover, if curtailment has taken 

 place, the ligament Ig^ must be an entirely new structure, which seems hardly likely. 

 I must repeat, however, that I have not seen Liitken's original paper, and am therefore 

 arguing against a view with which I am very imperfectly acquainted. 



The mode of articulation of the interspinous bones and fin-rays is decidedly interesting. 

 The interspinous bones are so arranged (fig. 25) that the posterior arm of one comes in 

 close contact with the anterior arm of its successor, the two being bound together by 

 ligament, and their ends sloped towards the point of contact so as to form a sort of shallow 

 cavity. In this cavity is fastened by ligament an ovidal nodule of cartilage (figs. 4, 20, 

 and 25, c.n), about \ inch in long diameter, upon which is perched, by its saddle-shaped 

 proximal end, a dermal fin-ray {fr). I have not met with cartilages of this kind in 

 any fish which has come under my notice, and can find no account of any such in the 

 works at my disposal ; I regard them as representing a second or distal series of radials 

 ox ])terygio]}hores^ , the interspinous bones forming the proximal series. 



The dermal fin-rays (figs. 4, 20 and 25, fr) are delicate, semitransparent, unjointed 

 bones, split longitudinally and vertically along their whole length into right and left 

 halves (fig. 4). Their proximal ends are expanded and pedate, and furnish saddle- 

 shaped surfaces, which articulate with the cartilaginous nodules described above. This 



' I propose to use the word pteryywpliore to signif j* any radial fin-supportiiig cartilage or bone in either the 

 median or the paired fins ; it wUl include, therefore, the cartilaginous fin-raj-s of Cyclostomes, Elasmobranehs, 

 and Ganoids, and the interspinous bones and brachial ossicles of Teleosts. In the latter group it will be 

 decidedly conTenient to have a single word which may be used instead of " interspinous bone," the awkwardness 

 of which in long descriptions is sufficiently obvious from the foregoing paragraphs. 



