^"i89i'"'] PROCEEDINGS OF THE NATIONAL MUSEUM. lf>9 



from the Murwnida', but uot by the characters thus giv^eii. As already 

 shown, the Murwnidw liave osseous brauchihyals and perfect skeletons, 

 probably, never have fewer than four opercular bones; finally, if 

 Myrocongcr belongs to the family, "a scapular arch" is at least some- 

 times present. It is in fact represented by cartilage in the typical 3Tu- 

 rcenid(v. 



Professors Jordan and Gilbert, in their "Synopsis of the Fishes of 

 North America" (1882), have recognized the family Murwnida; with 

 the same limits assigned to it by Cope. In their "Analysis of Families 

 of Apodes " (p. 35), they have contrasted the Murcenidw in a section {a) 

 characterized by ^'- preoperele wanting ; lower pharyngeals ic anting ; gill- 

 openings very small," with another section {h) distinguished by "pre- 

 opercle present; lower pharyngeals present." Inasmuch as no such 

 differences exist, the student would be at first thrown off the track in 

 his attempt to identify a murtenoid fish. The characters assigned 

 must not be considered, however, to be the results of independent 

 observations l)y the authors, for there is good reason to believe that they 

 solely relied upou Professor Cope for the osteological characters men- 

 tioned in their description of the family.* 



The limits of the family Murwnidw are well determined except in the 

 case of the genus Myroconger. That genus has been referred to the 

 Murwnina by Dr. Giinther simply because it is engyschistous. When 

 the extent to which fishes of other families vary in the extent of at 

 least the last branchial slit is remembered, the value of such a char- 

 acter may well be exaggerated. Nevertheless the genus Myroconger 

 may bo provisionally retained among the Murtenids until its osteology 

 or branchial appi\ratus is known. Inasmuch, however, as the total 

 suppression of the pectorals is characteristic of the typical Murwnidw^ 

 and as Myroconger has well-developed pectorals " about as long as the 

 snout," which itself is " of moderate length," it may be well to isolate 

 that genus as the representative of a peculiar subfamily {Myrocongrince) 

 and to keep it in abeyance as a doubtful constituent of the family 

 Murw7iidw. 



Not less than twenty-six generic names have been proposed for the 

 family. Many, if uot most, of these are undoubtedly superfluous, but 

 there is danger of going to an opposite extreme in reducing the number 

 to three or four (including Myroconger), as has been done by Dr. Giin- 

 ther. The course followed by Bleeker and Jordan seems to be the most 

 judicious, and thirteen genera appear to have characters entitling them 

 to such rank. A considerable range of variation is manifested by these 

 genera judging from the external appearance, may be found to be co- 

 ordinated with good osteological characteristics. 



'Professors Jordan aud Gilbert, in anotlier place (Syu. Fishes, N. A., p. 82), have 

 frankly acknowledged that "the osteological characters here [there] and elsewhere 

 in this work are mostly taken from Cope's Contribution to the Ichthyology of the 

 Lesser Antilles. Trans. Am. Phil. Soc, 1870." 



