22: MR, E. LEONARD GILL ON TUB PERMIAN 



The smaller fulcra are paired, but the larger ones at the base of 

 each series are single, the transition taking place through the 

 deepening of a median groove. The pectoral fin has on] 3^ three 

 (possibly sometimes four) fulcra, the distal half of its upper 

 border being unprotected (text-fig, 9). 



Of the fins themselves, two features are particularly note- 

 worthy ; the situation of the pectoral fin and the degree of 

 heterocercy of the caudal. The pectoral fin is placed rather high 

 on the fiank, and (at least in A. varians and A. alius) its lobe is 

 so formed that the fin can only be spread over or abreast of the 

 lower flank scales. It could not in life be spread below the level 

 of the body as in other fusiform Semionotids, with the pre-axial 

 border turned forwards and downwards. It is occasionally fos- 

 silised in this position, but it is then seen to be unnaturally forced 

 down so as to point forward over the jugular region or the bran- 

 chiostegal rays. 



The dorsal body-lobe of the tail is in a particularly interesting- 

 stage of reduction. It still extends to the extreme upper tip of 

 the tail, but for about the latter half of the distance it is repre- 

 sented by only a single line of small scales. Towards the tip 

 these scales are so small that but for their slightly different 

 shape they closely resemble the joints of the minute first dermal 

 ray lying beside them. (The appearance suggests the possibility 

 that on the withdrawal of the notochord these little scales might 

 be directly converted into the joints of an additional dermal 

 ray.) 



In other respects the fins greatly resemble those of Lepidotus. 

 As in that genus, a considerable length of the proximal part of 

 the rays is unjointed. In the best-preserved examples these 

 unjointed lengths and some of the joints immediately following 

 them show a flange-like lateral expansion, which led Kirkb}' to 

 state that they were covered by an accessory scale. The flanges 

 are especially conspicuous in the anal fin, where the rays are few 

 (only five or six behind the fulcra) and widely spaced. 



It is impossible to make out complete!}^ the bony structure of 

 the head, since none of the specimens show the elements scat- 

 tered, and all the more delicate parts are crushed together into 

 a fragile, semi-transparent plate no thiclter than fine notepaper. 

 Thus of the chondrocranium it can only be said that in the hinder 

 part it was considerably ossified, and that in one or two examples 

 a pair of exoccipitals are seen, and in another a club-shaped bone 

 in the position of the epiotic. A small bony rod occasionally 

 seen in front of the upper end of the pre-operculum may very 

 probably be the alisphenoid (text-fig. 7, asp.). 



Other internal bones of the skull are often shown, but never 

 in such a way that the}'^ can be completely described. The front 

 of the parasphenoid is very commonly seen as a straight rod 

 ci-ossing the orbit (text-figs. 6 and 13); it has a lateral flange 

 beginning a short distance behind its anterior end, but its hinder 

 portion remains unknown. The vomers are separate, in front 



