188G.] PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 81 



ueutly arrested and held at a certain point in the developmental scale of 

 Amiuriis withont any very obvious signs of concomitant atrophy. The 

 phylogenetic as well as ontogenetic modes of degeneracy are therefore 

 both obvious in the adipose dorsal of Amiuriis, but neither mode is more 

 than faintly evident in that of XoturuSj where the fin in question is 

 large, and i)ermanently, or almost perfectlj^, retains its embryonic pro- 

 portions. 



The median fin -folds of the lowest Chordata, viz, Branchiostoma 

 and Fctromyzon, are not su])ported by actinotrichia, but in the last, 

 especially, by dichotomous median cartilaginous rods wholly of meso- 

 blastic origin in the median tract. The Amphibia agree with them so 

 far as to have no actinotrichia, as far as known, in the median fin folds 

 of the larvoe, and are without cartilaginous supports for the same parts. 

 Whether the actinotrichia have degenerated in these last or not it is 

 now impossible to decide, because if they are totally wanting, as they 

 seem to be, it is now quite impossible to prove that they ever existed in 

 their ancestors from any evidence based on now living species, unless the 

 fossil remains of this type may have preserved evidences of their pres 

 ence in the older and presumably more fish-like forms. If the Amphibia 

 arose from some generalized type which gave rise to the fishes also, or 

 to the most generalized of the latter, then it would seem not unreason- 

 able to expect to find traces of the most primitive of all the types of 

 fin-rays, namely, actinotrichia, preserved in some of the Permian or 

 Carboniferous Amphibian remains. 



A structure may, however, be completely suppressed, and for so long 

 a period in some forms that their development will no longer recapitu- 

 late the comj^lete story of their i)hylogeny. This is illustrated for Phy- 

 sostomous as well as for Physoclistous forms in the genera Gamhusia 

 •A.\i& Hippocamims. Both of these last named genera have tended in 

 fact to revert very early to what I have called the archicercal stage, 

 the latter the most completely so of all known fishes except, perhaps, 

 Chima'ra monstrosa and Gastrostomus. 



AVith this I may conclude the i)resentatiou of the evidence in favor 

 of the use of embryological characters in the classification of fishes. 

 Such characters, it appears to me, may be used with just as much pro- 

 priety as any others; in fact with more, because the only possible way in 

 which the genesis or origin of any and all characters can ever be properly 

 understood is through a study of development. I have heard it stated 

 by systematists that embryological characters were of little or no value 

 in taxonomy ; in fact 1 once thought so myself, but upon a wider ac- 

 quaintance with the phenomena of development in certain groups I 

 believe I am warranted in saying that just in proportion as our knowl- 

 edge becomes more detailed and exact in reference to the small groups, 

 just in that proportion will we be able to avail ourselves of such char- 

 acters in taxonomy, and to appreciate exactly what is meant when we 

 speak of degeneration or specialization, 



Proc, il. M. 80 G Aii^-ii^t '^C, 1886. 



