74 FIN-RAYS IN CLASSIFICATION OF FISHES. 



tlie continuous fin which extends uninterruptedly along- the ba(;k, over 

 the tail, forward to the vent. No interradial spaces ever developed and 

 no further differentiation of the actinotrichia, except augmentation in 

 size and strength, and vastly more numerous than the serially arranged 

 actinophores or intersj)inous elements. The latter are cartilaginous, 

 with a membranons osseous investment, while the actinotrichia are 

 purely membrane and not cartilage, as asserted by Giinther. No sign 

 of atrophy of any part of the azygous fin-system is evident unless it may 

 be that there is an anterior portion of the dorsal and a i)reaual portion 

 which disappears. 



(U) The Iloloeephali, during development, probably approximate the 

 pre(;eding type up to a certain i)oint, when they diverge by ditlerentiat- 

 ing atrophied intervals, especially between the dorsal fin and the epaxial 

 l)art of the caudal, and in some species the tail-fold seems to atrophy 

 over the posterior portion of the chorda, so as to gi\'e rise to a nearly 

 cylindrical appendage without rays extending beyond the true tail, 

 and which may be called an opisthure. (See No. 1.*) 



(3) The next grade of differentiation of the fins is that seen in the 

 Elasmobianchs, in which, instead of there being a single row of actino- 

 trichia under the epidermis on either face of the fins, there are several 

 supei imposed. A tendency to form true permanent rays is also ap- 

 parent in large specimens, though the simpler Dipnoiin arrangement of 

 the actinotrichia is very apparent in very young specimens and in 

 cases where those have become rudimentary. There is also a less ob- 

 viously wide and continuous median fin-fold than in Teleostean em- 

 bryos, and a pronounced tendency' to differentiate a caudal, dorsal, 

 and anal fins, with intervening atrophied intervals between them. 



(It is very remarkable that Parapodoid structures should exist in the 

 embrxos of ScylliKm, as noted by Meyer (No. 7, pp. Iil9-U20), whovse fig- 

 ures also show that there is at first an archicercal terminal part of the 

 embryonic axis projecting beyond the point where median fin-folds are 

 developed, thus giving rise to a degenerate worm-like tail, such as has 

 been described by me as an opisthure (No. 1, p. 94). Such data as 

 jMeyer has presented are sufticiently conclusive, it seems to me, to add 

 great force to some of the conclusions reached by the writer in No. 5, pp. 

 1119-1121, but which were formulated still earlier upon other grounds 

 by Dohrn.t) 



The fourth grade of radial development in which several salient 

 charactei's appear for the first time is represented by the Chondros- 

 teans, Holostei, Crossopterygians, and Teleostei, which form a A^ery 

 natural group for other reasons, in like manner based on data which 

 embryological investigation has supplied. 



* The papers cited will be referred to by number. 



f8. Studien zur Urgeschichte des Wirbelthierktjrpers. VI. Die paarigen nnd un- 

 paaren Flosseu der Selacbier. Mittb. zool. Stat. Neapel, V, 1884, pp. 161-195, pis. 8, 9. 



