512 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 83 



attach themselves to seaweed or other floating substances, and are 

 liable to be carried by currents to great distances. Consequently, 

 some species are spread over different parts of the globe, like Anten- 

 narius, a genus the geograpliical distribution of which nearl}^ coin- 

 cides with that of Hippocampus." Though tliis may be true of the 

 geographical distribution of Antennarius or some of the species of 

 Hippocampus, it certainly does not apply to some of the species 

 discussed herein. On the contrary, two species of which a fair 

 amount of material is available, hudsonius and zosterae, are shown 

 here to tend to break up into distinct populations, which may be 

 separated by the ordinary statistical methods. These stocks are 

 found to occupy rather circumscribed geographical areas. Thus, 

 the populations of the subspecies hudsonius from Chesapeake Bay 

 and northward, from North and South Carolina, and from Mississippi 

 to Texas show average and statistically measurable differences. 

 This is also true regarding the populations of zosterae from Pensacola, 

 Captiva Pass, and Key West. (Tliis subject is discussed at greater 

 length under the respective species.) The tendency to break up into 

 distinct stocks within comparatively circumscribed geograpliical areas 

 is evidently a necessary consequence of the relative immobility of the 

 seahorse, which is a very feeble swimmer and probably spends its 

 life attached to seaweed in the vicinity where it was hatched. 

 Specimens of various species are sometimes pelagic and are often 

 taken in surface towdngs, but such definitely pelagic specunens as I 

 came across were immature or sexually undeveloped (p. 556). In 

 view of the fairly distinctive character of populations inhabiting 

 certain regions witliin the range of their species, it seems evident 

 that the few pelagic specimens that may be carried outside of their 

 range by waves and currents are not able to establish themselves, 

 grow to maturity, and reproduce in the region of their unmigration. 

 To the erroneous idea in regard to the geograpliical distribution of 

 seahorses evidently is to be ascribed, in part, the chaotic state in 

 wliich the systematics of Hippocampus is now found. Because of 

 the failure of authors in many cases to distinguish their specimens 

 properly, the geographical distribution of species and subspecies is 

 mostly unknown. The geograpliical range, as far as known or as 

 indicated by material examined, is presented under each species or 

 subspecies. 



UNCERTAIN SPECIMENS 



In an extensive study of variation among the species comprising a 

 genus, some extreme variants may be found that cannot be referred 

 to their proper species with certainty, WTiile such specimens usually 

 are comparatively few, they are of extreme interest and importance. 



