PHANEROSTEON MIRABILE. 167 



in 



ore or less crumpled, but it is very clear that the principal rays were articulated up to 

 their origins. From the same cause the shape of the ventral is also undefinable, but the 

 presence of a triangular-shaped pelvic bone is proved in more than one specimen. 



The dorsal fin is elevated and pointed, but has not exactly the same triangular- 

 acuminate form as in such typical genera as Elonichthys, etc. ; but that is simply due to 

 the longest rays which form the apex of the fin being further back, and those in front, 

 about eight in number, becoming gradually elongated. The exact number of rays 

 cannot be ascertained, but I think there cannot be less than nineteen or twenty. This 

 fin is supported on two sets of interspinous bones, proximal and distal. Those of the 

 proximal set (axonosts) are directly superimposed on the extremities of the neural 

 spines, but they are more numerous and consequently more closely placed ; their exact 

 number is not ascertainable. Their distal extremities articulate with the proximal ends 

 of the second set (baseosts), with which they correspond in number; the latter are 

 rather shorter and have both extremities somewhat dilated. 



The anal fin has the same form and size as the dorsal, and is supported by inter- 

 spinous bones of which I can only see one set. These are proportionally longer than 

 the corresponding elements of either of the upper sets, and are slender, enlarged at each 

 extremity, and diminish in length posteriorly, the last being only about half as long as 

 the first. As before it is difficult to determine the exact number of these ossicles. 



The caudal is not so deeply bifurcated as in the more typical representatives of the 

 family, or rather we may say that its lower lobe is proportionately not so long. It is, 

 however, very heterocercal and inequilobate, with a very strong upper lobe; its rays, 

 similar in general appearance to those of the dorsal, are, however, finely and minutely 

 dichotomised towards their extremities. 



The joints of the rays of all the fins are ganoid, and frequently show longitudinal 

 striations, but I have not, after most careful examination, been able to detect any fulcra 

 on the anterior fin-margins, unless we designate as such the median row of imbricating 

 V-scales which run along the upper margin of the caudal body-prolongation, as in other 

 Palaeoniscidse. 



Observations. — So many more or less perfect specimens of this most interesting little 

 fish have been now obtained from the Eskdale fish-beds, that there can be no doubt that 

 the above-described condition of the squamation is normal for the genus and species, and 

 that the condition in question is but one of the many forms of abnormal squamation 

 which we may see arising in this family of Palseoniscidae, in which the body is usually 

 completely covered witli the "regulation " array of rhombic and angular scales. Specia- 

 lisation has, in fact, produced many variations of form and condition of the scales in the 

 Palseoniscidae, such as — 



(1) The scales may become thin, rounded, deeply imbricating : Coccolepis, Agassiz ; 

 Thrissolepis, A. Fritsch ; Cryphiolepis, Traquair. 



(2) Body may become naked except for a few isolated patches of scales, e.g. behind 

 24 



