﻿24 



GANOID FISHES OF THE CARBONIFEROUS FORMATION. 



heemal arches with spines similar to those on the neural aspect of the notochord appears, 

 the semblance of vertebral bodies becoming now speedily lost. The arrangement of the 

 neural and haemal spines in the region of the caudal fin, or in the caudal body- 

 prolongation, is well shown in a tail of Nematoptychius Greenockii in my own collection. 

 Here a series of short bony pieces (neural spines) is seen immediately supporting the 

 row of V-scales, the so-called " fulcra," which runs along the ridge of the tail above. 

 On the inferior aspect of this caudal part of the body-axis, separated from the neural 

 spines by a space indicating the position of the notochord, and supporting the rays of 

 the caudal fin, is the series of haemal spines. Those supporting the lower lobe are of 

 considerable length and stoutness ; they get, however, very short as we pass back into the 

 upper lobe of the fin ; they are all expanded at their extremities, narrower in the middle, 

 the small ones of the upper lobe particularly tending to assume the hourglass form. 



The dorsal and anal fins are supported by well-developed interspinous bones. Those 

 of the dorsal are clearly in two series, one above the other ; there being first a row of very 

 short ossicles immediately supporting the rays, flattened and expanded at their 

 extremities, constricted in the middle, and diminishing gradually in size from before 

 backwards. They are often very well displayed in specimens of Rhabdolepis, contained 

 in the ironstone nodules of Saarbriicken and Lebach, and were described by Agassiz in 

 R. macropterus} Supporting these, and intervening between them and the neural spines, 

 is a second or deeper set, considerably longer and more slender in form ; their lower 

 extremities do not clip in between the neural spines, but seem simply to meet them. 

 There is no trace of such interspinous bones in front of the dorsal, as is the case in so 

 many other fishes, even in the by no means distantly related Platysomi. The inter- 

 spinous bones of the anal fin in Nematoptychius and Pygopterus are in front long and 

 stout, extending downwards and backwards from the first haemal spines ; they get 

 rapidly shorter and smaller posteriorly, so that an increasing space comes to intervene 

 between them and the last-named spines above, and in Pygopterus their direction 

 changes as we proceed backwards, so that the hinder ones come to incline downwards 

 and forwards at an angle to the direction of the rays. I can find no clear evidence of the 

 presence of two sets of interspinous bones in the anal. 



The Shoulder-girdle. 



Placed immediately behind the cranial buckler, there is, on each side, in Palceoniscus 

 (PI. I, fig. 5) and all the genera of the family, a plate (p. t.) of considerable size and 

 somewhat trigonal form, having its angles rounded and its surface gently convex externally, 

 and articulating with the bone s. cl. by its horizontal inferior margin. This I am disposed 

 to regard as the post-temporal (supra-scapular, Owen), and not merely as a nuchal plate, 



1 'Poissons Fossiles,' t. ii, pt. 1, p. 35. 



