J. K. TJiacher — Medhin (iiul Paired Fins. 291 



I m1s(. add tiouivs of the dorsal, I'l. LVIIF, fig. 01, and anal, PI. 

 LIX, lio-. 02, of A(-/j)enNer. Tlu'so arn very similar to the simpler 

 shark dorsals. It will be iioticed tliat we have the ])redoniinant 

 division into three pieces, hul llie terminal piece is very shoi't. 



(Joiiiiiisioiis rei/arditKj McxVihh Fins. 



'i'lie priniortlial median lin-rays in whatever form they oce\ir are 

 derivatives from a serit's of simple parallel chondroid rods, which 

 grew up in the me<lian fold in total independencte of the cartilaginous 

 arches alxni' and Itelow the notochord. These earliest representa- 

 tives of these parts weri' irom two to four times as numerous as the 

 vertebra' opposite them. In the (inathostomi true hyaline cartilage 

 replaced (he lowei- form of tissue seen in Myxlne and Petromyzon. 



Segnu'ntation and c(»ncresceiice, as well as redu(!tion in size, were 

 common changes in the (rnathostomes, and here a division into three 

 parts is the usual though not invariable rule. 



Hence it is seen that (iegenbaui"'s* st:itement that, in their sinijjlest 

 forni,^the prinioidial tin rays are mere pi-olongations of the neural 

 spines is incon-ect. It has been demonstrated that this was not the 

 earliest foi-m. The l)ij)noans, however, seem to oft'er an example 

 where the priim)rdial meclian fin-rays are mere prolongations of the 

 neui-al spines. They demand a moment's consideration. 



True neural spines are lirst found in the (ianoids. They are absent 

 in the Klasmobranchs and Agnathostomi. PI. LIX, fig. 0;i represents 

 the projection of a section of a vertebi-al segment of Acipenser cut 

 through the middle line of the arch ami neural spine. .\s the latter 

 slopes backwai'd, it is considerably fon'shortened in the figure. 



Now the cartilaginous arches spriiigiiig from the sheath of the 

 notochord pass upwards to lay themselves on each side of the fibrous 

 cord />, and here they sprea<l inward to meet oiu' another on the ventral 

 side of />, and also pi'olong themselves above to almost or quite meet, 

 and then they are followed by the dorsal spine a, from which they 

 are se])arated by a segmentation. 



Now the cord A, which is the same as the Uyainentum loiKjitadi- 



* Griiiidriss der Vergleichenden Anatomie, 1873, p. 488. Gregenbaur's assertion 

 in the .same place tliat they usually correspond in number to the vertebra? opposite, 

 is very strange. We have already seen that they do not do this in the earlier and 

 more significant forms ; and the statement of Gegenbaur would decidedly misrepresent 

 what we find, for e-xamplc, hi tlie figures of fish skeletons in Agassiz's PoisBons 

 Fossiles. 



