﻿88 



50 



second vertebra, il is in Centrisciis only partially covered by the cuirass and with 

 its upper margin firmly fastened to the end of the corresponding transverse process, 

 this besides being lodged in a flat pit on the inner face, as already mentioned. 



As to the scapular part I need only remark that the front end of the coracoid 

 (hypocoracoid Starks) remains unossified as in Amphisile, and that the foramen 

 of the scapula (hypercoracoid Starks) is enclosed by this bone alone, as in Amphisile. 

 Of the 4 basalia the lowest as in Amphisile is very large and here also 

 takes part in the boundary of an opening o, which however is more fissure-like '■''. 

 The number of fin- ray s in the pectorals I find to be most frequ- 

 ently 15, with in addition as in Amphisile a rudimentary ray on the upper border; 

 this is however somewhat more developed than in Amphisile. As in the latter the 

 rays decrease in size towards the lower edge of the fin ; 

 they are all articulated and unbranched. 



Pubic arch (figs. 30, 31). The characteristics 

 of the pubic arch are: (1) that its main part, i.e. that 

 which corresponds to the arch in other bony fishes, is 

 directed obliquely upwards and forwards (PI. II, fig. 2, i), 

 parallel with the lower end of the postclavicle which 

 lies immediately to the outside; in most other bony 

 fishes the pubic bones lie more horizontally in the 

 abdominal wall; (2) that a long horizontal process pro- 

 ceeds from the posterior end of the main part, longer 

 than the pubic arch itself (PI. II, fig. 2, p)\ this process 

 along with its fellow forms the roof of a groove into 

 which the ventral fins can be folded. 



If we compare this with the pubic arch in such 

 spinous- rayed fishes as Sebastes, Labrax, Trigla, Coitus 

 etc. it is easy to see that the essential features of the 

 arch in them are repeated in the anterior erect portion in Centrisciis; it is ex- 

 clusively on this portion that the muscles of the ventral fin rays are attached. 

 The thin, very prominent, lateral projection i corresponds to what Wintheh (35) 

 has called Crista superior in Trigla, Coitus etc.; the process e projecting vcntrally, 

 which bends inwards posteriorly and becomes suturally connected in the middle 

 line with its fellow of the opposite side is Winther's crista inferior. The arran- 

 gement of the muscles is in agreement with this (so far as I have been able to 

 determine from my badly preserved material); the dorsal muscles for the fin-rays 

 (i. e. Adductores and Abductores superiores, Winther, minus the Abductor for the 

 outermost ray) take up the space between the median suture of the pubic arch and 

 the crista superior i and pass through the hole b to the base of the fin-rays; the 

 lateral muscle (Abductor superior Wthr. for the outermost ray, the spinous ray) 

 occupies the space between the two cristæ /" and e; and finally the ventral muscles 

 (Adductores and Abductores inferiores Wthr.) occupy the space between the pubic 



Fig. 30. 

 Cciilr. scoIoptt.-i 

 I'ubic arch tron 

 above, i: erisl 

 superior; h: foi 



anien for tlie 

 tendons of Ih 



upper muscles 

 of (lie fin-rays 



p : liori/.onlul 

 process. 



Fig. 31. 

 Centr, scolopa.v. 

 I'ubic arch from 



below. Letters 

 as in fiK. 30. x: 

 cipeilins for the 

 teiiiliins of I lie 

 ventral muscles 

 of (he tin-rays: 



y: articulation 

 for the fin-ravs. 



