794 



SCIENCE. 



[X. S. Vol. XXI. Xo. 542. 



dark brown." Regan considers this variety 

 distinct from lima and names it garmani. Of 

 course, if the variety atropinnis is distinct 

 from the parent form it must go by the older 

 name atropinnis and not garmani. 



Eigenmann's genus, Loricaria, is divided 

 as follows: 



I. Teeth in the jaws in small or moderate number, 



not setiform; a more or less distinct orbital 

 notch. 



a. Snout rounded or pointed, not or not much, 



produced Loricaria. 



aa. Snout produced with a long rostrum, 



Hemiodontichthys. 



II. Teeth in the jaws numerous, setiform; orbit 



circular, without distinct notch. 



6. Dorsal opposite to the ventrals . OaJi/Zortcaria. 



Admitting that group I. is distinct from 

 group II. there is no reason why the second 

 group should not also be divided as group I. 

 is, and, indeed, Regan does divide it as fol- 

 lows, but does not use generic or subgeneric 

 names to designate the divisions : 



b. Snout rounded or pointed, not produced as a 



rostrum. 



66. Snout produced, forming a distinct rostrum ; 

 sides of the head in the male, margined 

 with bristles. 



The group under h contains the type of 

 Steindachner's genus Harttia; the second is 

 the Sturisoma of Swainson. 



The Hemiodontichthys of Regan contains 

 two distinct generic types, the one with the 

 snout expanded at the tip (the Hemiodontich- 

 thys of Bleeker), and the other with the snout 

 simply pointed (the Hemiodon of Bleeker). 

 Hemiodon being preoccupied, this genus may 

 be termed Beganella, in recognition of this 

 author's invaluable services in reviewing the 

 group. 



Arranging the respective genera or sub- 

 genera as far as possible opposite each other, 

 a comparison of the results of Eigenmann in 

 1890 and of Regan in 1904 gives us the fol- 

 lowing. 



Unfortunately there is nowhere in the paper 

 any indication what species were used as the 

 types of old or of new genera to help future 

 reviewers and similarly there is nowhere with 

 the synonymy and bibliography any indication 



Families and 

 Subfamilies. 



No. of 

 Species. 



Families and 

 .Subfamilies. 



No. of 

 Species. 



Genera. 





Genera. 





Subgenera. 





Subgenera. 





Argiidae. 





Argiinae. 





A rges, 

 Ci/i-hjjtiidn, 



41 

 2J 



1 . 



19 



A sirrjbtepiis, 



1 





1 



LoricariidEe. 





Loricariida. 





Lorieariinae. 





Loricariinae. 





Farloirellfi, 



6 





7 



Hemladonlichlh ys. 



1 





2 



Loricuria, 



30 





40 



Hartia, 



1 



\ Oxyloricuria 



10 



Oxyropsis, 







Hypoptopominae. 





>■ Hypoptopomatinae. 





Hypoptopoma, 







6 



Hiso'itoius, 









Parotocinclus, 



!i 







Otorinchis, 





1 Olocinclvs, 



9 



Plecostominae. 









Microtepidogaster, 













Neoplecostominae. 





Neoplecosiomws, 



2 



Ntoplecostomus, 



i 







Plecostominae. 





Plecostonws, 24'! Plecostoinus 



> ( Plecostomus 21 ) 



Rhinelepis, 4 ) {Pogonopoma 3) 



{Ehinekpis 2) 



Cochliodon, 11 „ 



Panaque, 3/ Panaque, 



Pseudacanthicus 

 Ancistrus, 



Hemiaiicisfnts, 18 ) {Lasiancistrvs 4) 



Pleryjoplichthys, 8 J {Ancistrus 18) 



26 



Parancistr 

 Pseudancistrus. 

 Delturus, 



[Parancistrus 3) 



(Psevdancisli-us 6) 



Hem ipsilichthys, 1 Hemipsilichthys, 



Acanthicus, 2 Acauthicus, 



Chcetostomiis, 20 ClicEtostomus, 



Ancistrus, 8 Xenocara, 



Total, 155 



1 



1 



15 

 12 



1S9 



of the locality from which species were re- 

 corded to help future students of geographical 

 distribution. 



The rules of nomenclature adopted differ 

 in principle from American usage. The first 

 species, the well-known Plecostomus plecos- 

 tomus Linnaeus, appears by the later name, 

 P. guacari Lacepede. Cochliodon Kner is re- 

 jected apparently on account of the use of 

 Cochlodon D'Orbigny, while Trichomycterus 

 is used, although it is a misspelling merely of 

 Thrichomycterus, which is another genus. 

 Cyclopium Swainson is rejected because ' his 

 generic name, being derived from the genitive 

 plural of Cyclops, is as inadmissible as would 

 be that of Silurorum.' Oxyloricaria is used 

 because the older Sturisoma is a ' nomen hy- 

 bridum.' The writer would be very glad to 

 be able to take back and make over the inele- 

 gant, barbarous or otherwise objectionable 

 names he has inflicted upon respectable fishes, 

 but, with his American confreres, he abides by 

 the rule, both for himself and for others, that 

 a name is a name no matter by whom con- 

 ferred or however wonderfully made. How 



