11)10.] CAUDAI, FIX OK rUE TKLICOSTOMI. 607 



of .s[)uciM.liz;itiou, p.articiilarly as re^anls t.Iie su[)prC'ssioM of tlio 

 virostyle ; but M,t the saiuo time the miiuberof liyj)m-al.s, viz, four, 

 associated with the hist vertebra is indicative of a comparatively 

 low type of .specialization. The pi-oximal ends of tlie lin-rays 

 embrace their eiidoskelebal supports so as to aliuost exclude them 

 from view, 



Ammodytes xobiancjs. (Ammodytida3.) (Plate XLIX. fig, 16,) 



The last vertebra and last hypural are completely fused anil 

 thus there is no free urostyle. The hypaxial supports consist of 

 live hypurals, four of which ai'e attaciied to the last vertebra. 

 The doi-sal elements are somewhat complicated : lying l)et\veeu 

 the single dorsal caudal radial and the last vertebra and hypural 

 are four separate ))ones, through three of which the spinal cord 

 passes ; these tlii-ee possibly represent neural arches, while the 

 fourth may be a dislodged arch, or a spine detached from its 

 arch. It is intei-esting to notice that fusion of these four bones 

 into a single piece would form a lai-ge neural arch exacth' as in 

 Jielone. 



AxHERiNA BAYERii. (Atherinid?e.) 



A small but free urostyle is found here ; connected with it and 

 tlie last centrum are five of the seven hypuial bones supporting 

 the hypaxial dermotrichia. A small cai-tilage, situated ))etween 

 the distal ends of the first two hypurals, shares in the su})port of 

 the anterior rays of the ventral lobe ; this may possiblj^ represent 

 a radial, but it is doubtful. Doi'sally the fin-rays are supported 

 by two dorsal caudal radials and one epural bone, the latttr 

 attached to the antepenultimate vertebra. The neural arch of the 

 last vertebra is very large and is composed of three bones ; it 

 alibrds protection for the end of the spinal cord. Thus AtlmruKc 

 ]iossesse.s a fairly generalized caudal fin, lacking only a ventral 

 c.iudal radial, tliough even tliis may possibly be represented l)y 

 the small cartilage mentioned. 



The development of the caudal fin in this fish is interesting as 

 being almost identical, in the earlier stages, with that of the 

 Flounder as figured by Agassiz, and shows very clearly tiie deep 

 incision between the original and the permanent caudals. 



MuGiL CAriTO, (Mugilid;\2.) 



The urostyle is lost in the tissue of the last hypural bone ; 

 three hypuiuls in all are attached to the last vertel)ral segment ; 

 three dorsal caudal radials are present, the last of which is api)lieil 

 to the dorsal side of the last hypural. The neural arch to tlie 

 penultimate vertebra is veiy much reduced. The anterior fringe 

 of the ventral lobe of the caudal is supported by small cartilages ; 

 what these lepresent it is difficult to determine. The fin there- 

 fore may be classed as an advanced tvi)e of homoceicv. 



40* 



