REVIEW OF HIPPOCAMPUS — GINSBURG 559 



Texas, has the caudal segments somewhat intermediate between the 

 two foregoing populations but nearer to that from North and South 

 Carolina, while the dorsal and pectoral ray counts diverge from the 

 northern population to an even greater extent than the population 

 from the Carolinas. The Gulf coast population also has a deeper 

 trunk and longer snout than the northern population. In all these 

 differences the two southern populations are intermediate between 

 the northern population of the subspecies hudsonius and the subspecies 

 pundulatus. It is evident that we are dealing here with a species 

 consisting of at least five distinct populations, three of which may be 

 regarded as populations of one subspecies while the other two diverge 

 sufficiently to constitute distinct subspecies. Attention may here be 

 called to the discussion of the geographic distribution of the species 

 of Hippocampus (p. 511). 



Geographic distribution. — The foregoing account and a study of 

 tables 1,3, and 4 show that the change in the structural characters 

 is gradual with respect to latitude. Consequently, it is evident that 

 geographically as well as morphologically an arbitrary line must be 

 drawn between the subspecies hudsonius and pundulatus. While the 

 most suitable boundary will need to be determined by a study of more 

 fish from intermediate points, it seems not far fetched to assign tenta- 

 tively those west of Florida as far as the Rio Grande on the Gulf 

 coast, and those north of Florida on the Atlantic coast, to the sub- 

 species hudsonius and those from Florida and Cuba to the subspecies 

 punctulatus. An inspection of tables 1,3, and 4 shows that on the 

 whole fish from North and South Carolina and from Mississippi to 

 Texas approach in their structural characters northern seahorses more 

 than those from Florida and Cuba. Consequently, the geographical 

 limits proposed are not altogether arbitrary but are based to a certain 

 extent on morphology. The arbitrary limit suggested would also 

 agree approximately with the general zoogeographical distribution of 

 the boreal and tropic piscine faunas in the western Atlantic. 



Material studied. — Off Seguin, Maine; October 1881; schooner 

 Charles Haskell (38900). St. Georges Banks; G. Brown Goode 

 (13110). Narragansett Bay, R. L; August 13, 1880 (25792). 

 Newport Harbor, R. I.; September 1, 1880 (26040). Off" Block 

 Island, R. I.; August 3, 1880; schooner W. M. Gofney (38950). 

 Patchogue, Long Island, N. Y.; September 14, 1884 (3G087). Off 

 Long Island; lat. 40°0r N., long. 68°54' W.; surface (31876). 

 Somers Point, N. J.; September 13, 1887; T. H. Bean (45102). 

 Great Egg Harbor Bay, N. J., August 23, 1887; T. H. Bean (45103). 

 Ocean City, N. J.; August 1, 1887; T. H. Bean (45104). Cliinco- 

 teague, Va.; July 1913; J. B. Henderson (76979). Off Virginia, lat. 

 37°27' N., long. 73°33' W.; surface; October 26, 1886; Albatross 

 (38189). Cape Charles City, Va.; October 1, 1897 (67885). Cape 



