1008 EEPOKT OF COMMISSIONER OF FISH AND FISHERIES. [28] 



Supplementary evidence of the occurrence of serial concrescence of 

 the radial supports may be found in the median fins of other fishes, as 

 it is very strongly marked in the dorsal and anal of the adult of Mola, 

 in which both neural and hsemal spines and interspinous elements are 

 crowded together and pushed forward posteriorly and backward ante- 

 riorly so as to condense these fins in an anteroposterior direction, so 

 that their bases are little more than half as wide as they would be had 

 no such distal approximation of their supports occurred. In conse- 

 quence of such concrescences the primitively-continuous median fins 

 are shortened, as in this case, or interrupted in other cases, in the same 

 manner as the lateral ones have been by the concrescence of their an- 

 terior and posterior segmental elements into a pectoral and ventral fin. 

 This fact does not, of course, discountenance the actual abortion of 

 some of the segmental radial elements which so frequently occurs, as 

 any one knows who has noticed the wide eradiate intervals between the 

 caudal and dorsal and the anal and caudal in the skeletons of numer- 

 ous si)ecies of Teleosts. 



This same kind of proximal concrescence occurs in the tail of Sal- 

 monoids, according to the investigations of Lotz (PL III, Fig. 4), where 

 certain hyi)nral elements originally distinct areBhown to be in the act 

 of coalescence. While it is true that in many cases actual concrescence 

 of the hyi)ural elements does not occur until ossification has been es- 

 tablished in the course of nearly completed development, I believe that 

 the origin in i)art of the diverging system of hypural bones and radial 

 ■I'lcments of the caudal of heterocercal Teleosts is to be rationally ac- 

 counted for on the basis of this principle, just as the diverging system 

 of rays of the paired fins is obviously to be similarly explained. It may 

 be that in the case of the Teleosts the urostyle, which is included by 

 the soft tissues of the tail, may be exserted beyond the hy])ural bones 

 for this reason, as seen in Salmonoids (PI. Ill, Figs. 1-4) ; that in fact 

 the more i)osterior hypural elements, as they are successively devel- 

 oped from chondrifications which take place in the median skeletoge- 

 uous tract, are shoved closer together basally than distally, so as to 

 leave a longer portion of the end of the chorda projecting than would 

 have projected had the bases of the hypural [neces been enabled to 

 maintain their normal position in the straight condition of the chorda. 

 This, I think, will be evident to any one who will take the trouble to 

 compare the stages of development of the Salmon (PI. II, Fig. 3) with 

 the condition of the tail of the adult (PI. VI, Fig. 2), copied from Lotz.* 



According to this view the notch marking the distinction between the 

 dorsal and ventral lobe of fish larvae in the act of becoming heterocer- 

 cal acquires a new significance. The epiblast is, in fact, shoved in at 

 this point somewhat in the same way as the post-pectoral epiblast is 

 shoved in at the hind part of the i^ectoral fold of Elasmobranchs, so as 



* Uebcr den Bau der Schwanzwirbelsiiule der Salmoniden, Cyprinoiden, Percoiden 

 und Ciitapbracten. Zeitscbr. wias. Zool., XIY, 1864, iip. 20, Pis. X-XII. 



