PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 



formulate the minor differences to be observed between fish of the s^e 

 species on different parts of our coast. A precisely parallel ca-e is to 

 be found in the shad of the different Atlantic rivers, which are well 

 known to exhibit strong distinctive marks. Very possibly every school 

 of menhaden has its own characteristics. In every case where I Lave 

 had an opportunity to observe them, the individuals composing the same 

 school were closely similar to each other. 



The typical form of the species as now defined is taken from the coast 

 of Southern New England ami the Middle States. It has the height of 

 the body about one-third of the total length, the head three-tenths of 

 the total length, or a little more, the maxillary long 0.14 to 0.14,3; and 

 exceeding the height of the dorsal. 



The species described by Spix under the name of Clvpanodon aureus 

 cannot be distingnished by any apparent specific characters from Bre- 

 cooriia tyrannus* since one or more of the specimens of the latter specie* 

 before me partake of some of the peculiarities of the Brazilian form. 

 There is, however, a general average of characters exhibited by the Brazil- 

 ian specimens as well as by the figure of Spix. with which they closely 

 agree, which seems to me to entitle them, for the present at least, to recog- 

 nition as belonging to a distinct geographical variety. The distinct: r«s 

 characters appear to consist in (1) a greater average height of body ; 

 (2) a lesser length of head : (3) a lesser average length of maxillary and 

 mandible; (4) a slightly lower anal and dorsal fin : o) a greater average 

 distance of anal from snout : (6) a greater average length of the middle 

 caudal rays ; (7) a shorter average pectoral ; (3) a more regular arrange- 

 ment of the scales, and a more luxuriant growth of small scales at the 

 bases of the fins. 



A number of specimens from Xoank, taken in 1S74. vary quite as mucii 

 from the normal type and in almost the same respect as the variety jusc 

 described. The maxillary and mandible are shorter, however, than in 

 the Brazilian form, the anal fin lower, and the lobes of the caudal are 

 extremely short, sometimes hardly exceeding in length the pectoral 

 fin. But for the fact that these specimens show almost all the charac- 

 ters of the Braziliau Breroortia, and in some cases exaggerations of 

 them, I should be inclined to consider the aurea a distinct species. 

 Having with some hesitation allowed it the rank of a variety, the ques- 

 tion must be decided as to the propriety of also allowiug varietal rank 

 to this pecaliar form from Xoank. The exact meaning of the terms sub- 

 species and variety as recently employed by zoologists is not very clear 

 to my mind, but I infer that a " variety" is composed of an assemblage 

 of individuals varying uniformly from the typical specific form in a 

 degree sufficient to be susceptible of description and definition, though 

 not necessarily separated from it by the absence of connecting forms. 

 Premising then that in giving to the Xoank specimens a varietal name 

 my object is simply to define the limits of variation from the normal 

 Proc, Nat. Mas. 7S 3 ?fny v l**7v 



