""^m"'] PROCEEDINGS OF TUK NATIONAL Mi;.Si:L'M. 71") 



torioiis work iucomplete in that he did not do no, and must accordingly 

 be especially blamed therefor. It may be hoped tliat the accomi)liNii('il 

 naturalist now in charge of the itollection (Prof. Leon Vaillant) will soon 

 Lave the work done, or, still better, do it himself. 



IV. SYSTEMATIC RELATIONS. 



In 1884 I expressed the hope that naturalists might '• make use ot' 

 their reasoning powers in considering [the three families of Tetrodon 

 toidea then recognized] and not <(s'si(me that they are nnjustidable be- 

 cause previous students had not appreciated their value." The hopf, 

 however, has not yet been realized. President Jordan, from whose 

 vigorous and progressive intellect most might havi> lu-en anticipated, re- 

 duced the families to subfamily rank, and thus reverted to my system 

 of 1878. A few remarks seem to be called for in defense antl support 

 of my later views. 



Such families as the Percidce, Serranidce, Pristipomidce, Sjxuidd^ and 

 Sqiiamipinnefi, with varying limits, are recognized by almost all ichthyol- 

 ogists of the present tin)e. Now what are the ditlcretjces between 

 them, compared with those between the three families into which the 

 Tetraodontoideans have been distributed? They depend, a^ generally 

 defined^ on slight differences in the dentition, armature of tlie opercular 

 bones, and extension of the scales on tins. Even if we look into the 

 internal structure, no very salient ditt'erences are observable; we l>e- 

 come, from such an examination, convinced that the characters that 

 have been generally used to difl'erentiate the families are abnost worth- 

 less as expressions of real alliiiities, but there is a striking general 

 resemblance in the crania and in all other parts, (.'ontrast with such 

 characters the differences exhibited by the crania of representati%'es of 

 the three families of Teiraodontoldea. The dilliculty is, tlieii. not to 

 ascertain the difj'erences, but to appreciate the resemblances. Many anat- 

 omists wouhl fail at lirst glance to understand the comparative homol- 

 ogies of the bones exhibited bv the several tvi)esof Tetraodontoideans 

 if they commenced their examination without any ])rcvious information. 

 And yet, forsooth, the families of Acantliopterygian lishes are generally 

 admitted, while one eminent ichthyologist unit«'S all the Tetraodontoi- 

 deans in a sinfile genus! Could inconsistency go farther f 



The differences between one of these tetraodontoidean families [Chon- 

 erhinida) and the others extend to the vertebral column in a marked 

 degree: most have a gieatly diminished number (tf vertebra- ; there- 

 fore the order, to accommodate them, has been said to have "the ver- 

 tebra' in small numl>er;"* the Chnurrhinids have more vertebra- than 

 a large proportion of the acanthoi)terygian tishea and certainly «lo not 



^Giiuther, Inf. Stii.ly Fishes, p. 683, 1880. 



