4G6 



EDWIN S. GOODRICH. 



Crossopterygii . 

 OstcolepidfE 

 Glyptopomidae 

 Illiizodortt.idse 

 Holoptychiidae 

 Crelacaiitliid.T 



Dipnoi 

 Conclusion 



PACK 



499 

 499 

 500 

 500 

 501 

 502 

 503 

 512 



Introduction. 



In the following paper are given the results of a sLud}', 

 carried on for some years, on the structure and development 

 of the. dermal fin-rays of fisb. The classification of living 

 and extinct fisli is a subject which has grown much in 

 importance of late years, and has been greatly advanced in 

 quite modern times through the labours of Cope, Traquair, 

 A. Smith Woodward, and others. Almost every available 

 character has been made use of in turn by systematists in 

 the endeavour to classify the larger groups ; and of these cha- 

 racters it is obvious that those which are based on structures 

 capable of being fossilised must be the most useful. No 

 parts of a fish are, as a rule, better preserved than the 

 dermal skeleton. Yet, although Agassiz (1) long ago 

 classified fish into large divisions according to the structure 

 of the scales, modern systematists are inclined to attach so 

 little importance to the exo-skeleton that I venture to think 

 they have somewhat neglected the dermal fin-rays. If it is 

 true that the application of cut and dried definitions of the 

 scaling will no longer enable us satisfactorily to subdivide 

 the Pisces into Placoidei, Ganoidoi, Cycloidei, and Ctenoidei ; 

 if it is true that '^placoid" scales may be present in the 

 Ganoidei, and that various forms of cycloid scales mny liave 

 been independently evolved from rhomboid ganoid scales in 

 later groups, — nevertheless these facts do not by any means 

 prove that the scaling is of little systematic importance. 



