SCIENCE. 



325 



natural arrangement — i.e. one based on their anatomical 

 details — would require, first, the fusion of the Giintherian 

 genera Centropristis, Anthias, Callanthias, Serranus, 

 Anyperodon, Prionodes, Plectropoma, and Trachypoma \ 

 then the wide removal of certain forms, and finally the 

 disintegration of the conglomeration on an entirely differ- 

 ent bas's from that accepted by Giinther. 



The instances wherein genera are referred to famines 

 with the diagnoses of which they diametrically disagree 

 are numerous. Leaving out of consideration cases of 

 conflict of genera or species with the characters assigned 

 as ordinal to the including group (e. g., Pogonias, Scice- 

 na, Gerres) the following are examples : 



The genus Dactyloscopus \s referred to the family Blen- 

 niida;, in which the spinous portion of the dorsal fin is 

 said to be " as much developed as the soft, or more." 

 Dactyloscopus has in the most evident manner, not- 

 withstanding the erroneous definition of Giinther (Cata- 

 logue of the Fishes in the British Museum, Vol. III., p. 

 279), only the first ten to twelve dorsal rays spinous, all 

 the others being articulated. In fact, Dactyloscopus has 

 nothing whatever to do with the Blenniidas, but is very 

 closely related to Leptoscopus, and belongs unquestiona- 

 bly to the same group ; in other words to an entirely 

 different division of fishes in the Giintherian system. 

 (See Trachinida; p. 462.) 



The genus Zoarces, (p. 497) also referred to the family 

 of Blenniida;, still more disagrees with the true repre- 

 sentatives of that family in the structure of the dorsal fin 

 and, as he himself admits, has " no other fin spines " than 

 a few near the caudal ; it shows, in fact, an organization 

 similar to that manifested in the family Lycodidas of 

 Gunther, (p. 537) placed by him in a different order of 

 fishes — the Anacanthini. 



Siphonognathus is a remarkable genus referred to the 

 family of Labridae. This family is defined as having, in 

 addition to other characters, " the soft anal similar to 

 the soft dorsa', ventral fins thoracic, with one spine and 

 five soft rays," and " branchiostegals five or six." Noth- 

 ing whatever is said respecting the anal, ventrals, or 

 branchiostegals. of Siphonognathus and as the necessary 

 data are thus entirely suppressed, it would naturally be 

 assumed that the genus would' have the characters at- 

 tributed to the family. In fact, however, Siphonogna- 

 thus has not the "soft anal similar to the soft dorsal," 

 there are 7/0 ventral fins, and there are only /owrbranchios- 

 tegal rays. It will be thus apparent that it would be im- 

 possible to identify this fish from Dr. Giinther's Intro- 

 duction unless it were assumed that great blunders 

 had been made. This is indeed the case, but it is 

 not safe to assume that the author is an habitual 

 blunderei , and to proceed on that basis, even in the 

 case of Dr. Gunther. We are somewhat prepared, how- 

 ever, for the idiosyncrasy exhibited by Dr. Gunther, 

 when he compares the relationship of Siphonognathus to 

 Odax as being similar to that of Babirussa to Sus (see 

 Catalogue of Fishes in B. M., v. 4, p. 243). Any one who 

 can really entertain such views, and consider the differ- 

 ences between the mammalian genera to be of the same 

 kind or degree as those between the fish genera is unfit 

 to institute comparisons. 



Numerous genera are adapted, which, although they 

 may be good, consistency would require Dr. Gunther to 

 merge with others. Thus we have Ptyonotus (which he 

 has unnecessarily substituted for Triglopsis of Girard) 

 retained for a form in the family of Cottidas (p. 480) ; this 

 is, however, far more closely related to the " Codies quad- 

 ricornis " of Gunther than are any of his other species 

 of that heterogeneous group. Pamtnelas is still retained 

 as the name of a distinct genus which is allied to Trachy- 

 notus, although it had been named before Dr. Gunther 

 applied his, and its affinities have been well-known for 

 many years to be with Ccntrolophus : it is indeed to a 

 species of that genus (the C ovalis), that the P. perci- 

 formis is most closely related, and yet in spite of the con- 



current testimony of previous ichthyologists we find it in- 

 jected, in the " Introduction to the Study of Fishes," into 

 a family remote from that to which Centrolophus has been 

 referred. As examples of other forms unnaturally sepa- 

 rated we may instance (1) Chatopterus (p. 390) and 

 Aprion (p. 397) ; (2) Grystes (p. 392) and Huro (p. 393), 

 and (3) Auliscops and Aulorhynchus (p. 508). The last 

 type, it may be remarked, is more nearly related to the 

 so-called Gasterosteus spinachia than to the Fistu- 

 lariidae and should be either referred to the same family 

 or differentiated as a distinct one. 



Changes of the names of established genera on trivial 

 pretexts are also indulged in. The name of Triglopsis 

 was abandoned for Ptyonotus because there was a Triglops 

 previously established. Although they are unquestion- 

 ably much alike, they are sufficiently different, and Stein- 

 dachner has even lately named a genus Atherinops, know- 

 ing well that Alherinopsis had already been proposed for 

 another genus of the same family. Dactylopus is dis- 

 carded for Vulsus because, forsooth, the term Dacty- 

 lopoda had previously been applied by Meyer to a 

 group (not genus) of extinct reptiles. And yet our author 

 himself retains both Chondrosteus and Chondrostei 

 etc., without the slightest demur. Xiphasia is rejected 

 with an exclamation mark (!) and the yet more objection- 

 able name Xiphogadus proposed because the author was 

 dissatisfied with the name, and — we strongly suspect — still 

 more with the namer (Swainson). Why expect any better 

 reason ? 



The idea is conveyed in the work — and that it has been 

 extensively claimed elsewise by our author is no secret — 

 that all the established genera are admitted in this vol- 

 ume. Without counting the scores of genera that Dr. 

 Gunther refuses to recognize, but which every one ap- 

 plying the canons observed by mammalogists and ornith- 

 ologists would adopt, there are many which even that 

 author could scarcely neglect unless through ignorance. 

 Among those omitted, and which are especially interest- 

 ing, on account of representing previously unknown 

 types of high value (families or sub-families), or because 

 they throw light on the relations of families in which 

 they belong are : Elassoma, Xenichtkys, Hoplopaqrus, 

 Gnathanacanthus, Nematistius, Grammicolepis, Bathy- 

 tnaster, Cottunculus, Oxylebius, Anoplopoma, Dactylag- 

 ntts, Myxodagnus, Anarrhichthys, Plagiotreinus, Cha- 

 nopsis, Nematocentrzs and Protistius. If he had really 

 known Hoplopagrus (referred to incidentally on page 

 279, but not otherwise noticed), he, perhaps, would not 

 have so far separated his " Percida: " (pp. 375-379) and 

 " Sparidos " (405-410), as he has done : if he had known 

 Cottunculus he would, perhaps, have recognized the 

 affinity of Psychrolutes to the Cottida, and not isolated 

 it as the type of a remote family — at least no scientific 

 ichthyologist would have (ailed to so profit by the know- 

 ledge. The work of Bleeker, Steindachner, Klunzinger, 

 Lutken, Vaillant, Sauvage, Giglioli and Collett in Eu- 

 rope, and that of all American ichthyologists has, how- 

 ever, been almost of nought so far as Dr. Gunther is con- 

 cerned. It need be only remarked, in connection with 

 the latter, that of the numerous genera of Etheostomine 

 fishes only Pileoma (Percina) and Boleosoma (p. 379) are 

 recognized. The reason therefore is no secret — they are 

 too small, and as they have not been able to grow larger, 

 they do not deserve to be considered. The interesting 

 relations, physiological and morphological, that they pre- 

 sent are not sufficient to outweigh this cogent objection. 

 Among American fishes there is no group that has been 

 so much written about and that is better known than the 

 genus Micropterus, but notwithstanding Dr. Gunther 

 has not yet learned that he has distributed its well defined 

 representatives under three genera, nor that Huro was 

 based on a mistake and is not a valid genus, nor that there 

 are two, and only two, well-determined species, and those 

 two can not be generically distinguished. When it is 

 further remarked that only three genera are recognized 



