1042 REPORT OF COMMISSIONER OF T1>U AXD FiSHERTES. [62] 



oibcr i)arts of the skeleton well ossified. The embryonic condition 

 of tile fin niys of Ceraiodns is in j)erfect Ueeping with the develop oent 

 of tlie rest ot the animal's skeleton, which is very largely cart iagi 

 nous, so that in a general way it closely resembles in the degre of 

 its (leveloj)n)ent that trantiiory stage of the Salmon euibryo '^•heu 

 the embryonic rays are still distinct and parallel with each other. To 

 find such a i)arallelism existing between an embryo Teleost auc the 

 adult of l>ipyioi is, to say the least, very suggestive of the thought that 

 the Teleostean and Dipnoan phyla are remotely affiliated. On no <tlicr 

 ground are wa enabled to understand why it is that a Teleost should re- 

 capitulate so closely in the course of its development conditions which 

 are permanent in the Dipnoi. 



Protopicrus resembles Cerafodun, according to Wiederslieim,* in the 

 structure of its median fin-vSystem, and presents the same dorso-ven- 

 tral symmetry of the U])per as contrasted with the lower half of the tail. 

 The main difference which ihese forms present when compared with the 

 lophocercal condition of the caudal end of the body of fish larvse is 

 the j)resence of partly osseous neural and hiemal spines, interneural, 

 interha^mal, ami basilar interneural and basilar interhamal elements, 

 which. support numerous rays nearly equivalent to the embryonic fiu- 

 rays. 



The tail of the Crossopterygian Folypterus exhibits a tendency to be- 

 come het»rocercal, hardly, however in proportion to the extent to which 

 ossification has proceeded throughouttheentire column. The persistency 

 ol the axial symmetry of the caudal fin we must therefore regard not nec- 

 essarily as s\ud>olic.il of its degeneracy or completed evolution, but 

 rather of the persistence of conditions which have not distnrl)ed that 

 symmetry. That the condition of dorso ventral symmetry found to ob- 

 tain at tlie posterior part of the chordal axis of fish embryos generally is 

 the most ancient is (.-onclusively shown by the evidence « erived liom 

 all known tyi)es of fish larvre. 



As already indicat* d by competent observers, the loi)hocercal condi- 

 tion ]>recedes the heterocercal, which is itself followed by the outwardly 

 homocercal condition ; yet there are instances known in which this rule 

 is violated to some extent, as I have already pointed out elsewhere. 

 These are where (l) the median fins are not developed from a continu 

 ous median fold, as in Siphostoma, Hippocampus, and Oamhusia ; and (2) 

 where a truly lophocercal homocercy precedes the structurally hetero- 

 cercal stage of development, as in Alosa, for example. In both of these 

 instances, however, there is no reason to believe that we have any era- 

 bryological principles contradicted, but that in reality those principles 

 are confirmed by these apparent exceptions. 



It is therefore noteworthy that the primitive embryonic rays in the tail 

 of the embryo of Alosa have a ])erfectly symmetrical disposition above 

 and below the caudal axis, as shown in the dotted lines in Fig. 2, PI. 11. 



* Yergleichendea Auatomie tier Wirbekhiere, Jena, 1883 and 1884, 



