REVIEW OF HIPPOCAMPUS — GINSBURG 569 



to the Cuban population of that subspecies; in the depth of its trunk, 

 the length of the snout, and the pectoral ray count it is nearest to the 

 southern populations of hudsonius, while in the dorsal count it is 

 nearest to the northern population of that subspecies. Although the 

 number of specimens from Bermuda studied are few and the precise 

 range of variation of this population remains to be worked out, it 

 seems apparent that if hudsonius and punctulatus are to be recognized 

 as subspecies, kincaidi also should be recognized as having equal 

 rank with them. 



In its comparatively lower tubercles, fewer caudal segments, and 

 slenderer body kincaidi approaches reidi, and the differences between 

 them are discussed under the latter (p. 575). 



Material studied and geographic distribution. — Bermuda (23795, 

 F. M. Hamlin, 1879; 23805, G. Brown Goode, 1877; also Field Mus. 

 Nat. Hist. nos. 5064, 5065, 5066, and 5495, T. H. Bean). 



Total number of specimens examined, 6; 4, with a brood pouch, 

 75 to 118 mm long; 2, -without any trace of a brood pouch, 61 and 62 

 mm. Apparently kincaidi is now known only from the coast of 

 Bermuda. 



Nomenclature and synonymy. — ^ Although the types of kincaidi and 

 brunneus were not examined, they unquestionably pertain to the 

 subspecies here described. Apparently the former was based chiefly 

 on the strongly developed tubercles and their long, branched fila- 

 ments, while brunneus was based chiefly on color, the presence of 

 large blotches in the form of hourglasses. The present study definitely 

 determined that in hudsonius, as well as in punctulatus, the high 

 tubercles, the profuse filaments, and the blotches are normally 

 juvenile characters that often persist in medium-sized or even nearly 

 full-grown specimens (see pp. 51 1 and 555). Evidently the same varia- 

 tion occurs in kincaidi, although I do not have sufficient specimens 

 to determine this definitely. The tubercles and filaments of kincaidi 

 indicated on the published figure and the color of brunneus as described 

 show that neither was based on specimens of reidi, the other large 

 seahorse occurring at Bermuda. 



Both kincaidi and brunneus were established on misapprehensions, 

 since the characters that apparently induced their describers to estab- 

 lish the names are well shown by the subspecies hudsonius and punctu- 

 latus during certain stages of growth or as an individual variation. 

 However, since the Bermuda population is subspecifically distinct 

 from hudsonius and punctulatus on the basis of other dift'erences, the 

 names kincaidi and brunneus, the former having priority, are available 

 for that population. 



