546 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 83 



anticipated by Schinz and possibly by two other previous authors. 

 The name longirostris Schinz, 1822, certainly, and possibly ramulosus 

 Leach, 1814, and rosaceus Risso, 1826, if the last two were based on 

 specimens of the same species, have priority over guttulatus Cuvier, 

 1829. According to the strict application of the rules the later name 

 guttulatus should be suppressed. Nevertheless, since it has become 

 so well estabhshed general usage is here followed and the name 

 guttulatus continued. This course is more expedient also for two 

 reasons: (1) The proper application of the name ramulosus, which 

 has priority over longirostris, must remain uncertain until the type 

 is reexamined, and (2) a name earher even than ramulosus may be 

 discovered as applying to the form. While I attempted to examine 

 and review all the early publications bearing on the nomenclature of 

 the seahorses, it is quite possible that I missed some pertinent publi- 

 cation. As a matter of fact, I came across Schinz 's name longirostris, 

 which has been left out of all general lists, by mere chance. Further 

 search may reveal a still earlier designation for this species, which 

 would necessitate another change of name. Therefore, the use of 

 the well-established name guttulatus is continued. 



As stated previousl}^, my study has shown that the populations of 

 this species from the Atlantic and from the Mediterranean coasts 

 are subspecificaliy distinct. Consequently, it becomes necessary to 

 restrict further the application of the early names, and the name 

 longirostris Schinz, as well as the later substitute name guttulatus, has 

 been formally restricted to the Mediterranean population (see p. 525). 



Uncertain specimen. — A single specimen in bad condition from 

 Greece (45041) probablj^ belongs to this subspecies. The dorsal 

 and pectoral are injured and the number of rays cannot be accuratel}^ 

 determined. Trunk segments 11, caudal segments 38, the two upper 

 ridges overlapping on two segments; without a brood pouch; length 

 91 mm; depth 13, snout 11, postorbital 11, head 25.3, trunk 32.7, 

 tail 63, and orbit 5.3 percent of length. From table 2 it niaj^ be noted 

 that these measurements agree fairly well with guttulatus except in the 

 unusually long snout. This may represent an extreme variant in 

 that respect, or it may have some taxonomic significance, a question 

 to be determined only by a study of more numerous specimens from 

 Greece. It is possible that guttulatus is divisible into distinct popula- 

 tions, as are hudsonius or zosterae (see under their accounts). 



HIPPOCAMPUS EUROPAEUS Ginsburg 

 FlGUKE 58 



Hippocampus brevirostris Yarrell (not Schinz, 1822), A liisloiT of BritiHli fishes, 

 vol. 2, p. 452, 1836 (England; the rather short snout shown by the figure 

 indicates that the account is probably based on the present species). 



Hippocampus europaeus Ginsburg, Journ. Washington Acad. Sci., vol. 23, p. 561, 

 1933 (La Rochelle). 



