"^"wol"'] proceedings of the national museum. Ill 



<0(fe8, genera intermediate between the northern Schastcs and the troj)ical and subtrop- 

 ical representatives of the family of Scor2)nnid(v, was also intermediate. 



But while claiming the generalization that there is a correlation between the increase 

 of vertebra'- and the increase of latitude among fishes, T would not assign it an undue 

 value or claim for it the dignity of a law. It is simply the expression of a fact which 

 has no cause for its being now known. It may be added that this generalization is 

 true only in a general sense. 



Jordan, 1801. — In another paper* tbe present writer bas said: 



This increase in the number of vertebra' in northern forms has been used as a basis 

 of classification of the Fleuronectidw by Jordan and Goss, of the Scorj^a'nida; by Jordan 

 and Gilbert, and it will doubtless iirove to have a high A^alue in the subdivision of 

 other families which have representatives in difierent zones. The cause of this pecul- 

 iarity of fishes of cold waters is still obscure. Probablj' the reduction in number of 

 segments is a result of the specialization of structure incident to the sharper compe- 

 tition of the tropical Avaters, where the outside conditions of life are very favorable 

 for fishes, but the struggle of species against species is most severe. 



In this paper is given a table which shows that in the genera of La- 

 hrid(v\ inhabiting northern Europe and the New England waters there 

 are 38 to 41 vertebra?, in the Mediterranean forms | 30 to 33, in certain 

 subtropical genera § 27 to 29, while in those Lahro'uU which chiefly 

 abound about the coral reefs || the number is from 23 to 25. 



Jordan & Eigenmann^ 1891.^ — In a recent paper on the Serranidw{^Qii- 

 bass and groupers) it is stated that the grouj) as a whole belongs to the 

 tropical seas, and that it differs from the related fresh-water family of 

 Percidm by the muclr smaller number of vertebra?, nsually 24, which is the 

 number most common among spiny rayed fishes. Amowgiha Serranidw, 

 however, two genera form exceptions to the general rule. One of 

 these,** with 35 vertebra'., occurs in the rivers of China, the other,tt 

 with 36 vertebra?, in the mountain streams of Chili and Patagonia. In 

 these two genera the numbers are materially increased, as would be ex- 

 pected if the rule is to hold good. There are, however, other Serranidce, 

 more or less perfectly confined to the fresh waters, and yet retaining the 

 normal number of vertebrae. These are perhaps comparatively recent 

 immigrants from the sea. In evidence of this is the fact that among 

 these forms there is a perfect gradation in habits from the strictly ma- 

 rine,f| through migratory§§ and brackish-water speciesHH to those con- 

 fined to the rivers and lakes.^^f 



So far as I know, the above record includes all the references to this 

 subject yet made in ichthyological literature. We may now examine 

 the facts in detail. 



* Review of the Labroid Fishes of America and Europe, p. 2. 

 \Labrus, Acantholabrus, Ctenolahrus, Hiatula. 

 t Chiefly belonging to Sym2}hodus. ^ Lachnolaimus, Harpe, etc. 

 II Scarus, Sparisoma, Xyrichihys, Julis, Thalassoma, etc. 



H A Review of the Genera and Species of Serranidw found in the waters of America 

 and Europe, by David 8. Jordan aud Carl H. Eigenmanu. 

 ** Lateolahrax. tt Fereich th;/s. 



U Dicentrarchits punctatus. ^^ Eoccus Jineatus. 



111! Morone americana. HH Boccus chrysops. 



