Cockerell: Scales of Characinid Fishes. 93 



tions are that the neotropical region is their original home. Africa 

 appears to have been supplied with only a few types, perhaps three or 

 four, at long intervals of time. Whether one or two may have gone 

 by a southern route in very early times, it is now impossible to say; 

 but I do not see why the ancestors of the Hydrocyoninae might not 

 have arrived via Asia, during the Tertiary, at some period when the 

 northern climate was warm and America and Asia were continuous. 

 The period of mmigration into North America and Asia (probably 

 two or three genera only) might have been relatively short, and the 

 chances of finding any fossil remains might therefore be very remote. 

 The close resemblance of the Erythrinine scale to that of Old World 

 rather than New World Cyprinids must surely be significant. 



Anodin^. 

 Scales not seen. 



CURIMATIN.E. 



Curimatella alburnus (Miiller and Troschel). Scales broad, semi- 

 circular in form, the base strongly pleated and wavy; two more or 

 less imperfect apical radii, far apart; circuli moderately densely 

 transverse in the apical field. This is a quite ordinary Curimatine 

 type; the apical margin is not at all dentate. 



Psectrogaster and Curimatus are figured {cf. Plates XXIII-XXV). 

 The scales in this group are broad, the base approximately straight 

 except for the strong crenulations. There are often two forms of 

 scales on the sides of the same fish, one with the circuli dense, the other 

 with them much less so. The apical radii are variable, but there are 

 often a pair of strong ones (Aphyocharax type), and others weak, or 

 rudimentary. The margin in the scales seen by me is not properly 

 ctenoid, but merely inclined to be toothed between the weak radii. 

 Curimatus, Curimatopsis, and Psectrogaster do not essentially differ 

 in the scales. The weaker type of scale, as in some (immature?) 

 specimens of Psectrogaster, is like the African Citharinus. Citharinus 

 has only the weak system of radii. Gill describes Psectrogaster auratus 

 Gill, from Bolivia, as having the scales all deeply pectinate. The 

 figure of the scales of P. ciliatus (M. & T.) now given shows only wavy 

 irregular teeth, but the specimen is probably immature. The teeth, 

 even if well-developed, would have no resemblance to those of Dis- 

 tichodus, etc., but may be compared with those of Citharidium. 

 The following species have been studied: 



