KEVIEW OF HIPPOCAMPUS — GINSBURG 527 



KEY TO THE SUBGENERA AND THE AMERICAN AND EUROPEAN SPECIES OF 

 HIPPOCAMPUS -^ 



. Dorsal rays 16 to 31. Pectoral rays 13 or more. Upper ridges of trunk and 

 tail usually overlapping on two or three segments, infrequently on one 

 (in hudsonms as an individual variation) . First caudal segment hexangular, 

 infrequently quadrangular as an individual variation. Base of dorsal over 

 3 to 6 segments, usually including first caudal segment. 

 6>. Dorsal rays 26 to 31. Caudal segments 44 to 49. Upper ridges of tail and 

 trunk usually overlapping on three segments. Dorsal usually over 6 seg- 

 ments. Trunk segments 12 or 13-. Subgenus MACLEAYINA (p. 529) 

 ¥. Dorsal raj^s 16 to 21. Caudal segments 33 to 40. Upper ridges of tail 

 and trunk usually overlapping on two segments, infrequently on one or 

 three as a a individual variatior (with exception of ingens about as often 

 on three as on two). Dorsal usually over 3 segments, sometimes partly 



or wholly on a fourth segment Subgenus HIPPOCAMPUS (p. 530) 



c^ Trunk segments normally 11, sometimes 12, rarely 10 as an individual 

 variation (10 segments in one specimen of hudsonius out of entire 

 number studied) . 

 d^. Tubercles on upper ridge either well developed and more or less 

 pointed or at least narrowly rounded above, or else nearly obsoles- 

 cent, not in the f-arm of broad and low stumps, 

 e'. Tubercles on upper ridge comparatively well developed and con- 

 spicuous, at least in specimens up to about 150 mm long (except 

 usually obsolescent on trunk in large males of punctulatus and 

 Jcincaidi less than 150 mm long). 

 /I. Trunk without dark transverse lines or large blotches; white dots 

 on side of trunk numerous. Northern Mediterranean, eastern 

 Atlantic, and eastern Pacific coasts. 

 g^. Snout in medium-sized females (118 mm or less) long, more 

 than 10 percent of length; relatively long also in males when 

 like sizes are compared; trunk comparatively slenderer when 

 like sizes are compared (see table 2). Whitish dots often 

 very profuse, minute, and subequal all over, tending to form 

 very fine white streaks. Profusely covered with small dark 

 spots. Penultimate trunk segments about as often novem- 

 angular as septangular (slightly oftener novemangular in the 

 specimens examined). Attains to a large maximum size. 



Pacific coast of North and South America ingens (p. 534) 



g'^. Snout not more tlian 9.9 percent of length in both sexes in 

 medium and large specimens. Trunk averages deeper. 

 Whitish dots usually not so profuse, coarser on trunk and 

 head, often coalescent there to form short irregular bands or 



3' The purpose of this key is twofold: (1) To give a synopsis of the most important specific characters in 

 concise form, and (2) to facilitate the identification of specimens. The student is warned, however, not to 

 expect to be able to "run down" specimens in every case by the use of this key. It is impossible to con- 

 struct such an ideal key for the species of Hippocampus. One important drawback to the construction of 

 such a key in this genus is the necessity of using the structure of the tubercles for specific distinctions. 

 While the differences may be appreciated readily when specimens are directly compared, it is impossible 

 to convey in descriptive phrases an adequate picture of these diflerences. Moreover, the structure of the 

 tubercles differs considerably with size and sex in the same species, and human language is not gradated 

 finely enough to express these diflerences and their variation, except in general terms. This key, therefore, 

 may be used to full advantage only in connection with authentic specimens for comparison. However, 

 at least full-grown or medium-sized fish may be identified by the use of this key, together with the tables 

 giving the frequency distributions of the meristic characters and the ranges of proportional measurements, 

 and with a knowledge of the locality of capture of the specimens to be identified. 



