324 



SCIENCE. 



ters, and was, in some respects, a step backwards, al- 

 though of not very much moment. 



Another and most radical modification — the next stage — 

 maybe fitly noticed in the author's own words, [i.] 

 "The discovery (in the year 1 87 r ) of a living representative 

 of a genus hitherto believed to be long extinct, Ceratodus, 

 threw a new light on the "affinities of fishes. [2.] The 

 author who had the good fortune of examining this fish, 

 was enabled to show that, on the one hand, [3] it was a 

 form most closely allied to Lepidosiren ; on the 

 other, that it could not be separated from the 

 Ganoid fishes, and therefore that also [4] Lepi- 

 dosiren was a Ganoid : a relation pointed out already 

 by Huxley in a previous paper on 'Devonian Fishes,' [5] 

 This discovery led to further considerations of the relative 

 characters of Miiller's sub-classes, and to the system 

 which followed in the present work" (pp. 25-26). In regard 

 to this claim there are several noteworthy and character- 

 istic features. 



(1) In 1870, in Dr. Giinther's Cat. Fishes Brit. Mus., 

 vol. 8, p. 323, it is expressly admitted that " after [the 

 ' sheet ' descriptive of Protopterus and Lepidosiren} had 

 passed through the press, Mr. Krefft informed me of the 

 most interesting discovery that a living representative of 

 Ceratodus had been found in Queensland. Nothing 

 of this genus was hitherto known beyond teeth, as those 

 described and figured by Agassiz in Poiss Foss. iii, p. 129, 

 pis. 18-20." (2) Dr. Giinther knew nothing whatever of 

 Ceratodus till he received a communication respecting it 

 from Mr. Krefft. (3) As indicated by Dr. Giinther himself 

 (Trans. Royal Soc, v. 161, for 1871), Mr. Kreff', in even 

 the title of his paper, published April 28, 1870, and before 

 Dr. Giinther's "reply had time to reach Mr. Krefft," recog- 

 nized the affinity of the genus to Lepidosiren. (4) As 

 early as i860, Gill (as Brandt, Peters, Liitken and others 

 subsequently recognized) showed that " Lepidosiren was 

 a Ganoid," and that Polypterus was a type intermediate 

 between the ordinary Ganoids and the Dipnoi. (5) Conse- 

 quently the only novelty in Dr. Giinther's work was " the 

 system which is followed in the present volume," which 

 has been pronounced by an eminently competent judge to 

 be " a triumph of systematic gattcherie." Whatever is 

 true in the statements examined had been appreciated be- 

 fore Dr. Giinther labored and only what is untrue to 

 nature and to science was original with him. The co-or- 

 dination of the facts enumerated was the necessary logical 

 result of the successive steps. 



But what is " the system which is followed in the present 

 work?" Only the salient features may be noticed, and 

 these will sufficiently appear from the enumeration of the 

 sub-ordinal, ordinal and super-ordinal groups. These 

 are : 



I. Sub-Class— Pal^ichthyes. 



I. Order — Chondropterygii. 



I. Sub-order — Plagiostomata [Sharks and Rays]. 



II. Sub-order — Holocephala [Chimaeroids]. 



II. Order — Ganoidei. 



L Sub-order — Placodermi [Extinct]. 



II. " — Acanthodini [Extinct], 



III. " —Dipnoi. 



IV. " — Chondrostei. 



V. " — Polypteroidei. 



VI. " — Pycnodontoidei [Extinct]. 



VII. " — Lepidosteoidei. 



VIII. " — Amioidei. 



II. Sub-Class — Teleostei. 



I. Order — Acanthopterygii. 



II. " — Acanthopterygii Pharyngognathi. 



III. " — Anacanthini. 



IV. " ■ — Physostomi. 



V. " — Lophobranchii. 



VI. " — Plectognathi. 



III. Sub-Class — Cyclostomata. 



IV. Sub-Class — Leptocardii. 



To those familiar with the facts and details of the an- 

 atomy of fishes and the inferior vertebrates, this enu- 

 meration will be its own best commentary. Suffice it for 

 the present at least to affirm that it involves more con- 

 tradictions and inconsistencies than have been mani- 

 fested in any recent taxonomical exposition of any class 

 of animals emanating from a respectable source. 



Almost equally in disaccord with the cultivators of the 

 other branches of Vertebrate Zoology is Dr. Giinther in 

 his treatment of Genera. 



The extreme of differentiation is practiced by ornithol- 

 ogists, (provided the differences are obvious and external), 

 and a course is pursued in mammalogy which has received 

 the sanction of the greatest number of students of that 

 class, during at least the last quarter of a century. 

 American ichthyologists have endeavored to comply 

 with the principles on which genera in the latter class 

 have been recognized as much as the differences of facts 

 will permit, and although, of course, there are many dis- 

 agreements as to detail, there is an essential congruity 

 between them. The principles, if any, applied by Dr. 

 Giinther are undiscernable from his work. His methods 

 indeed, seem to have varied with the whim of the mo- 

 ment and to have been modified for each case : the re- 

 sults then happening appear for him to have crystallized 

 and not to have been subject to review or further consider- 

 ation afterwards. Strange contrasts constantly occur in 

 the extension or limitation of the groups. In the genus 

 Tetrodon, for example, is discoverable a very consider- 

 able range of variation, not only in external features but 

 still more markedly in the details of structure, and espec- 

 ially in the bones of the head. So great are these that 

 there are three well defined major groups and a number 

 of minor ones entitled to generic distinction, but, never- 

 theless, our author has refused to admit more than one 

 "genus " for all the representatives of the type, whereas, 

 in the related group of Diodontines, he has recognized a 

 number of genera upon characters of very much less 

 moment, such as the development of the spines, nostrils, 

 &c. Under the genus Gasterosteus are confounded all 

 the representatives of the family of Gasterosteids, and 

 yet upon differences of the same kind as those which 

 distinguish, for example, the " Gasterosteus spinachia " 

 from the other species of Gasterosteus, are elsewhere 

 constituted distinct families. 



These examples might be extended indefinitely. Heter- 

 ogeneous combinations of forms on one hand chance in 

 strange contrast with isolated generic types on the other. 



Comprehensiveness of genera per se is not a great 

 evil, provided there is consistency in the treatment of the 

 subject, and that all share as nearly alike as the nature of 

 the case allows. It is to the assignment of inordinate 

 value to a few superficial characters, and the subordina- 

 tion, to the manifestation of such, of other characters 

 whose coincidence demonstrates them to be of greater 

 importance, that we object. It is true that the accept- 

 ance of such comprehensive groups isolates in a measure 

 the class in which they are recognized from others 

 and tends to constantly mislead the inquirer who would 

 compare the constituents of the several classes, e.g., as 

 to their geographical or geological relations. Even this, 

 however, is of minor importance. It is the utter disie- 

 gard of the gradations of structural differences exhibited 

 by Dr. Giinther in his constitution of genera that detracts 

 so much from the value of his work. To enter into de- 

 tail would necessitate space equal to the portion consid- 

 ered, and some instances must suffice. 



Serranus (p. 381) is distinguished among its allies by 

 the " small scales," presence of " very distinct canines in 

 both jaws," and the absence of serratures from the lower 

 margin of the preoperculum. Under the genus thus de- 

 fined, there are not only species which disagree with the 

 principal characters, but the typical Serrani (S. cabrilla, 

 S. scriba, etc.) are more nearly related to the species of 

 Centropristis than to the rest of their associates. A 



