^^^89l!''] PROCEEDINGS OF THE NATIONAL Ml'SKl'M. 709 



wards, liowever, Lave shown good reason for not doing so longer, ;iml 

 also reason not good. Tlioir own words may be reproduced : 



Tlie proper name of the geiiu.s has been involved in Hoine confiiHion, owin^ to tlie 

 use by Swaiuson of two ditVerent names for it. 



On page IIM of Ills miserable work on the classificaticHi (»f fiHhes, SwuinHon j^iven 

 an analytical key to the genera, and applies to the present gronp the nan)o of Canthi- 

 ijusti'v (correi'tly written .icanthofitisttr). No species are here nientionc<l Jiy Swainson, 

 but in this case his diagnosis is accnrato and snllicieiit. On page :W"^, these genera 

 are again defined, the present one in nearly tbo same way, but nnder the name of 

 rsihinotiin. Two species (roKlratas: clectricun) are here mentioned as types. 



I'rofessor Oill has jirefcrred to adopt the last-mentioned name, reganling ( (inlhi- 

 tjanteras nnideiititiabitM-xecpt tlirongh the mc^diiim of the Hjieeies mentioned under the 

 diagnosis of Pailonottts. Dr. Bleeker has preferred to take theearliername of Canthi- 

 ffaster. In this case it is certainly trues that no doubt could exist as to wliat Swain- 

 son intended to include nnder ('aiifhiijaster, even hail the second iliagnosis been 

 omitted ; moreover, the name I'xilonotna is preoccupied. We see, therefore, no sufll- 

 cient reason forsetting this name aside, objectionable as it is. 



lean only account for Jordan and Ed ward.s's assertion that Swainson's 

 diagnosis of Cantliiga.ster is " sullicient" by the surmise that they 

 have considered the diagnosis of Psilonotu}i instead of Canthujuster. The 

 sole diagnosis of Swainson's Canfhit/a.sterisui the word.s: ^' Canthitjaster : 

 Muzzle prolonged and narrow ; belly with spines." Now, if Messrs. 

 Jordan and Kdwards consider this sufficient, 1 do not, and 1 tlnd it as 

 applicable to some species of ^'tSpluvroifles,'^ especially S. angufiticeps, as 

 to the species recently referred to Psilonotm. Indeed, the resembhmce 

 of *S\ atu/Ksticcps to the Psilonoti is so great as to have misled two incom- 

 l)arably better ichthyologists than Swainson — Richardson and Stein- 

 dachner — who actually referred that species to the genus, the former 

 calling it Aiichi.somus au{)ustircps and the latter Caiithoijnsfrr lohntus.* 



Nevertheless, Jonlan and Kdwards were ([uit*' right in correcting me 

 for adoi)ting the name Psilonoiiis, but only because that name had 

 been preoccupied in llymenoptera, a fact of which 1 was not aware in 

 1884. 



Under the circumstances, however, it may be doubtful what name to 

 take up for the genus in (piestion. Cnnthiijustcr^ with its apparent ety- 

 mology, is a very objectionable name, and its api>lication, as already 

 urued, could m)t have been certaiidv determined "except through the 

 medium of the species mentioned under the name of I'sihniotHs.'" Hut 

 as it can be so determined, I am disposed, after IJIeeker, .budan, and 

 Edwards, and some others, tu adopt it. It would naturally Im" supposed 

 that the name was intended to allude to the spiniferotis belly, and 

 Swainson uiidoul)tedly labored under the delusion that " <7njf/ii "— or, 

 ''('rt><//<».s" wasa good Creek deiivative for spine. ♦ Swainson, however, 



•The external dilVerence between the species of Canthiijaiittr are marked l)Ut they 

 are not indicated by tlie words of Swains«)n. 



tSee Canthophnis (vol. 1, p. Sf.l ; vol. •.>, j.. :{ll)), CanthilrpUa (vol. -J. pp. 7. .V2. 179 

 2i\l),<!iUHa,itlu,s{vo\.-2, pp. 17(1, '2\'2), Pohjcanthun ( vol. 'i, pp. K.'i. 242), ele. Another 

 delusion was that leptes was a Cireek derivative for scale or scaled. 



