780 Btdletin 4y, Untied States Natiotial Museum. 



quality of the group. The presence of intrusive elements has been one 

 of the chief causes of the failure of schemes of classification in the past. 

 The modern fashion of a comparatively minute subdivision of genera and 

 families has this justification, that "Analysis must precede synthesis." 

 The large families and genera recognized by the earlier writers were 

 always disfigured by the retention of " aberrant," that is, unrelated forms. 

 (d«rtx'ftt, spine ; nrspuv, fin; the word originally written Acanthopterygii, 

 but the shortened form seems preferable.) 



Analysis of Suborders and other groups of Acanthopteri. 



Note. In the following analysis only the most salient or the most evi- 

 dent characters are mentioned, detailed descriptions being given farther 

 on. The groups called suborders are mostly 8hari)ly defined and more or 

 less isolated from the main trunk of the mass of spiny-rayed fishes. The 

 other groups mentioned below are for the most part incapable of precise 

 definition, representing rather centers of relationship. Various aberrant 

 forms receive provisional location pending more exact and detailed study. 

 Thus under the Scombroidtl, Berycoidei, and TrachinoUUi for example, are 

 here enumerated forms which may have little real affinity with the cen- 

 tral family of the group in (juestion. In this connection may be quoted 

 the following pertinent remarks of Dr. Gill (Mem. Nat. Acad. Sci., vi, 128, 

 1894): 



" The author insists, as in previous publications, on the entirely provi- 

 sional nature of the present arrangement. Changes — grave changes — must 

 necessarily be made in the system when the species shall be studied in a 

 more scientific way than has been generally done heretofore. Many fam- 

 ilies are entirely unknown in an anatomical point of view, and until their 

 structure has Ijeen investigated and carefully compared with that of 

 others, their systematic relations must remain doubtful. No scientific 

 investigator should fear to change his opinion. An obstinate persistence 

 in ancient views because they have been once adopted has been too long 

 detrimental to the interests of systematic ichthyology, and such obstinacy 

 has retarded the general progress of science for twenty to thirty years. 

 While the aspect of every other branch of vertebrate zoology has entirely 

 changed within that period, ichthyology, the most complex and the least 

 advanced of all, might appear to the casual observer to have had a more 

 certain basis than any, inasmuch as the text-books of a past generation 

 have essentially the same system as the latest. Either almost omnis- 

 cience and prescience were the attributes of the guides of the past who 

 keep to the same path in the present, or obstinacy and blindness to an 

 extraordinary degree have been manifested. To a great extent ichthyol- 

 ogy has been limited to descriptions of species or habits, and taxonomic 

 principles have been quite neglected. 



Meanwhile ichthyology is still a crude and inchoate science. The i)re8- 

 ent list has been prepared as a check-list and adjuvant to the use of the 

 collection. Many families have been allocated in tiieir positions simply 

 because they have been placed there before and because equal doubts 

 would be involved in placing them elsewhere. In many cases, it is cer- 

 tain that the general conceptions of their relations (if so positive a term 

 may be applied to what are vague reiterations of past utterances) are 

 baseless, but the indications furnished by the exterior are insufficient to 

 justify positive conclusions. Until the anatomy or at least the osteology 

 of every family and subfamily is known, much doubt must remain as to 

 the proper allocation of such groups. (Gill.) 



Salmopeuce; Adipose fin present; dorsal and anal with spines in very small number; ventral fins 

 abdominal, with more than 5 soft rays; vertebra; about 35. 



Xenarchi: Vent anterior in position; ventral fins thoracic, with more than 5 soft rays; dorsal 

 and anal spines few; tail diphycercal. 



