June 23, 1916] 



SCIENCE 



899 



An inherently stable machine striking a head 

 gust of J feet per second soars to altitude of 

 about 4i J feet above its initial level and, 

 after executing oscillations, remains about 3J 

 J feet above the original level. 

 9. Terms of Relationship and Social Organi- 

 zation: Truman Michelson, Bureau of 

 American Ethnology, Washington, D. C. 

 From the point of view of Algonquian tribes 

 terms of relationship are linguistic and dis- 

 seminative phenomena, though in other cases 

 they may be primarily psychological and socio- 

 logical. 



Report of the Annual Meeting: Prepared by 

 the Home Secretary. 

 This report has appeared in full in Science. 



Edwin Bidwell Wilson 

 Mass. Institute of Technology 



SPECIAL ARTICLES 



THE SCALES OF THE GONORHYNCHID FISHES 



The Gonorhynchidse constitute a small family 

 ■of very peculiar marine fishes of elongate form, 

 found in the seas about Japan, Australia and 

 South Africa. In the Eocene deposits of Wy- 

 oming is a fish which Cope named Notogoneus 

 osculus, considered to belong to the Gonorhyn- 

 chidse. Whitfield in 1890^ gave an account of 

 a specimen of this species, and expressed the 

 opinion that it belonged in the vicinity of the 

 suckers, or Catostomidee. It seemed remark- 

 able that a fish from a fresh or brackish water 

 deposit in Wyoming should be referred to a 

 rare marine family of a remote region of the 

 earth ; and the scales of Notogoneus, admirably 

 figured by Whitfield, did not at all resemble 

 those of the Isospondylous fishes in general, 

 neither had they any resemblance to those of 

 the Catostomidse. Wishing to apply the more 

 exact methods of comparison of later times, I 

 asked Dr. D. S. Jordan for scales of Oono- 

 rhynchus, and he has very kindly sent mate- 

 rial from 0. abhreviatus Schlegel, obtained by 

 Alan Owston in the Yokohama (Misaki) 

 market, Japan. These scales wholly confirm 

 the reference of Notogoneus to the Gonorhyn- 

 chidse, and aSord a remarkable illustration of 

 the constancy of scale-structure through miU- 



1 Bull. Amer. Mus. Nat. His., III., p. 117. 



ions of years and migrations over the earth. 

 The long parallel-sided scales of G. abhreviatus 

 are narrower than those of N. osculus, and the 

 truncate base is crenulate, but the peculiar 

 structure is entirely the same. The apical mar- 

 gin has a single row of 18 or fewer (never so 

 many as in N. osculus) teeth, which are long 

 and stout, and connected by a thin lamina. 

 Just below these is a broad sculptureless band, 

 the same in living and fossil forms. The 

 lateral circuli are strictly longitudinal and not 

 very dense. Spreading fan-like from the sub- 

 apical nucleus are the radii (about 12), 

 closely set, with longitudinal bands of curved 

 lines, derived from the system of circuli, be- 

 tween them. 



Jordan and Snyder^ say of G. abhreviatus: 



Mr. E. C. Starks has examined the shoulder 

 girdle of this species ; it has the mesoeoraeoid arch, 

 as usual with Isospoudylous fishes. Its place is 

 apparently with the earliest and most generalized 

 of these forms. 



The scales, however, are more like those of 

 Acanthopterygians. Coming to details of 

 structure, we find a striking resemblance to 

 the scales of Aphredoderus, of which genus 

 Jordan says : " Probably the most primitive of 

 all living Percoid fishes, showing affinities 

 with the Salmopercai " (to which group Eegan 

 has more recently referred it). Aphredoderus 

 has the same type of marginal teeth, though 

 there is no hyaline band beneath them and the 

 radii are few. Marginal teeth of the same 

 type are found in another group, little related 

 to Aphredoderus or Gonorhynchus; namely, 

 the Characiform genus Distichodus of the 

 fresh waters of tropical Africa. The rest of 

 the Distichodus scale shows no close resem- 

 blance to that of Gonorhynchus. 



We have, then, evidence of the extreme con- 

 stancy of scale characters, even minute de- 

 tails, in the Gonorhynchidse. On the other 

 hand, the most striking feature of the Gono- 

 rhynchid pattern appears, not in the presumed 

 allies of that family, but in other families 

 supposed to be very far removed from it. Is 

 this wholly a matter of independent evolution, 



2 Smithsonian Misc. CoU., 45 (1904), p. 236. 



