300 J. K. Thaoher — Median and Paired Fins. 



derived directly from the lateral folds of Amphioxus. These in their 

 turn may be referred to a reduplication of the process which has 

 already formed the atrial space, but which is not carried so far here, 

 in the lateral folds. 



Ventral F'm of Aci2')enser. 



The ventral fin of a young specimen of Acijyenser hreiiirostris is 

 exhibited in PI. LIX, figs. 64 and 05. The fin of one side is separate 

 from that of the other, no synchondrosis uniting the two halves of 

 the girdle. But the part J> approaches closely its fellow. The same 

 separateness of the two sides obtains in the shoulder girdle. The 

 composition of the fin is peifectly evident. Beginning at the aborad 

 end of the row, we have first three separate and parallel rays. The 

 proximal joints increase in length from the first to the last of the 

 three. In the remaining rays these basal joints, increasing still moi'e 

 in length, have united with each other to form the large pelvic piece 

 ab. The composite nature of this is confirmed by the groovings of 

 the surface, which extend about a centimeter before they finally fade 

 out. 



The iliac process, «, is half a centimeter high. 



In fig. 64 the proximal joint of the penultimate ray grows up a little 

 under the last ray, in the manner familiar in the median fins. It does 

 not happen to occur in the fin of the other side, fig. 65. 



The predominant three-fold division obtains. But the penultimate 

 ray in fig. 64 has a tip or an extra segmentation, and c is without 

 the distal segmentation. 



The breadth and the outline of c raises the suspicion of its double 

 character. 



I have had >io opportunity of examining other Ganoid fins, and 

 this one of Aci2^enser seems, on the whole, that which most nearly 

 approaches the parent form of the Gnathostomes. But while in the 

 independence of the two sides, in the separateness of the I'ays, and 

 the simple segmentation, it gives us the early form more complete 

 than is elsewhere found, in the number of rays and hi the absence of 

 the iliac process the shark ventrals are less advanced. 



Elasmohraiich YentraU. 



We now turn to the derivation of the ventral fins of sharks, one of 



the two abutments of the Inroad S{)an of the archiptcrygiuru theory. 



The series of steps by which I conceive them to have been derived 



