HEVIEW OF HIPPOCAMPUS — GINSBURG 557 



Preserved specimens often do not show the typical color pattern. 

 Some are very dark, the color pattern being then much obscured or 

 nearly obliterated, and some are very light all over, the pattern then 

 being very faint or nearly absent. Often specimens are irregularly 

 mottled without any definite color pattern. However, although not 

 alv/ays well marked and varying greatly with the individual, the 

 typical color (consisting of a blotched pattern in the young, partly or 

 wholly replaced by a striped pattern in large specimens) is charac- 

 teristic of hudsonius as well as its subspecies pundulatus and probably 

 also kincaidi. It was not observed in any of the specimens of the 

 other species studied, except the single specimen tentatively identified 

 as villosus (p. 582), wliich to some extent has the blotched appearance 

 of hudsonius, although not so well marked as in typical specimens of 

 the latter species. 



Distinctive characters and relationships. — The relation of the com- 

 mon large seahorse of the more northern States to the one from Cuba 

 and Florida apparently has never been definitely established, but it 

 becomes clear by referring to tables 1 and 3. After reviewing cur- 

 rent general works on American fishes one gets the idea that two 

 common species of seahorses, hudsonius and punctulatus, occur on 

 the Atlantic coast of the United States, the former ranging farther 

 north and the latter being more southern in its distribution. Accord- 

 ing to some authors ^^ both of these common species may be found 

 at the same locality. This assum.ption is certainly an error, as the 

 data presented herewith prove. Table 1 shows that fish from Chesa- 

 peake Bay as compared with those from Florida and Cuba average 

 more caudal segments, fewer pectoral rays, and fewer dorsal rajs 

 (not a greater number of dorsal rays, as erroneously stated in current 

 descriptions). As the proportional measurements of the different 

 parts of the fish differ with age and sex, no adequate picture of the 

 frequency distribution of these measurements could be shown by the 

 available material, but the ranges and the averages are given in 

 table 3. This shows that when large specimens of the same sex are 

 compared, northern fish, on the average, have a slenderer trunk and 

 a shorter head, shorter subdivisions of the head (snout and post- 

 orbital), slightly shorter trunk, and somewhat longer tail; although 

 the dift'erences in proportional measurements nearly disappear in 

 smaller fish. However, while tables 1 and 3 show distinct and sta- 

 tistically measurable dift'erences in the seahorse populations from 

 the extreme geographical ranges, they also show a high degree of 

 intergradation. Furthermore, this intergradation in the structural 

 characters is evidently gradual with geographic distribution or 

 latitude, and fish from North and South Carolina and from Missis- 



« Smith, The fishes of North Carolina, pp. 172-173, 1907; andJordan, Evermann.and Clark, Rep. U.S. 

 Comm. Fish, for 1928, p. 244, 1930. 



