,H9i. J PROCEKDlNCiS OF THE NATIONAL MIISKUM. 7<»7 



tieiinlUoch, not L.; was supposed liy Swainson lo l»o hucIi. If wo rfKin.l, with Dr. 

 Gill, thissuhilivisiou to Ins properly .i rostriclioii of the Liiiiiumii ;;i<mis, tlio iiaino 

 Tetraodon would again bn nynonymoii.s with Arotliron. Bnt it may l»o ohji-clod that 

 the Telrodoit of Ssvainson coiitaini'd no species known to Linn.iMiH, and henet' ilHcuni- 

 positiou can not bo considered as a prolan- restriction. This objection s.-ems to ns a 

 valid one. 



The next subdivision seems to be that of Miiller (IH41), who n<taine<l the name of 

 Tetrodon for nonc^ of his divisions. 



Next (Id-'ifj) we have the snhdivisioii of Hibron. liy liiia ihe jjronp wasdiviiU-d 

 into a larj^o number of genera, part of them without delinition and all of them with 

 French names only. For one of his sections the name Tetraodon was rclatued. This 

 group, as arranged by Hibron, included a single Linna>an species as type. This one, 

 liinatns, is a member of the group called by Miiller Arotliron. This seems to be tli<- lirst 

 proper rostrictioa of Tetrodon, and, so far as wo can see, it must stand, making Tetra- 

 odon the equivalent of .Irolhron. 



Later, 1^37, lloUard worked over the material of Hibron, and ailopted— on skele- 

 tal characters only— au arrangement of genera, not unlike that given in the prc«<!nt 

 paper. His genera are Xenoptcnii (not American), RhifnchottiH (=f'anthiganler), 

 IhtlrachojiH ( = Colonir'<iin). nrachiicephalus ( -Tetraodon), Apuiccphnhis {=zSphirroidr9 

 and L(iiiorcphaliiH), Monotreta (not American). 



llollard supplies a Latin form to the French names of Hibron, and using the name 

 Titroilon as a general term, ho places Hibron's Tetraodon as a subgenus uuiler liis own 

 Briichi/vephalux. 



In 1S()7, Hleeker. i)robably regarding lincattts as the projier type of Tetraodon, 

 seems to have su[»pressed the latter name as a syiumym of the uame Craijracion , used 

 by Klein before the date of the Systema Nat iu:e. Other ichthyologists ilo not give 

 Klein's names pre<-edenee over those of Liniueiis. an<l under tin- rules ol' nomi-ncia- 

 tnre which we adopt, Cratjracion must l»o disregarded.* 



In H7.), Prof. Gill used the name Tetrodon as synonymous with Lagucrphalnt, 

 and in 188') as synonymous with Arotliron. In lrt"<3, Jordan and (Jilbert regar.ied 

 T. tentiidintuff as its type, thus making it synonymous with Sjiliirroidrn. 



It seems evident to us, from the above data, that it is best to re-^ixnl Tetraodon 

 UncdtiiH ns thoty\}e o( Tetraodon, &nd thus make the latter ninie the eiuivalent of 

 Arotliron. 



The rallacy in tliis iirgiiimMit is in coiisidci iiig tin- iiiiiiic Tttrumlnn in 

 the stiine li^j^ht as a new generic name. Swainson tli»l not pivteml to 

 eiuinieriite all the species of the genus. He simply .selected some, •>! 

 which there existed Mgnres in ti cotiple of illiistr.itcil works accessililc 

 to himself. Inasmuch as those so selected were congeneric with Lin- 

 luean species, those Linn;e;in si)ecies were l)y implit ation inclnil»'(l and 

 actually do belong to the genus tis litnitiMl by Swainson. Of course, if 

 Swainson had given ;i new geiuM'it! name, the name cottld only have 

 been rettiined for siiecies iictutilly included by him under the genus 

 named; its it wtis, he simply limited a gt'uus already «'stablished, and 

 the genus so limited indudetl, by the terms of its »liagiu)sis, tw(» of the 

 five original species of Tctraodini, \\\\\W not more tli.m one Linna-an 

 species belongs to any other genus. 



1 revert, with Jordan and Kdwards, to the name Tetraodon for this 



* Tetraodon iin<\. Craijracion were both useil by Hleeker. the former being applied 

 to the genus Lagocephalnn and the latter being essjMitially eninvalent to .Irothron of 

 Miiller, or Tetraodon of Jordan and Edwards and American authors. 



