236 PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 



and If iu iuterorbital width ; no dark vertical half-bars noticeable on 

 the sides of these specimens. These specimens are therefore interme- 

 diate between Mollienesia latipinna and Mollienesia Uneolata, as charac- 

 terized by Jordan & Gilbert (Proc. U. S. Nat. Miis., V, 1882, 258-260). 

 Doubtless these differences noted are mere local variations in speci- 

 mens from different waters. 



4. Heterandria formosa Agassiz. 



Head 3^ to 3§ in length ; depth, 8^ to 4 ; D., 7 ; A., 8 or 9 ; scales 

 in lateral line, 28. Body short, slightly compressed. Snout very 

 short, about two-tliirds eye. Eye large, 3 in head. Mouth terminal, 

 slightly oblique. Lower jaw slightly projecting. Jaws each with a 

 series of small, pointed, movable teeth. Gill membranes united. Dor- 

 sal fin short, on posterior part of body ; its origin above, on a vertical, 

 from middle of anal, and about midway between end of snout and 

 tip of caudal fin. In male specimens the anal is considerably in ad- 

 vance of dorsal, and is transformed into an intromittent organ. Caudal 

 fin rather long, about 5 in body, slightly dusky at its tip. 



Color in spirits brownish-olive. A dark band about as wide as eye 

 extends from mouth through eye and along middle of the side, termi- 

 nating in a black spot at base of caudal. This band is crossed 

 with from 6 to 9 brownish-black vertical streaks which become fainter 

 with age, the anterior ones the less prominent. These markings are 

 made up of small dark dots ; black spots on base of dorsal and anal fins. 



Description from about 30 specimens, both sexes represented, taken 

 from a small brook near Jacksonville. The largest is 0.9 inch in 

 length. The genus Heterandria was proposed by Agassiz (Amer. 

 Journ. Sci. Arts, XVI, 135) in 1853 to include certain Cyprinodonts 

 which have the anal fin in the male modified. The two species men- 

 tioned, lioJhrooki and formosa, belong to different genera, which corre- 

 spond respectively to Poey's Gambusia and Girardinus. These occur 

 in the Memorias sohre la Historia Natural de la Ida de Cuba (I, p. 390). 

 The date assigned to their appearance by Girard (Proc. Ac. Nat. Sci., 

 Phila., 1859-'61) is the year 1851, which would give both two years pri- 

 ority over Heterandria. Giinther and Jordan «& Gilbert have ac- 

 cepted this date without challenge. It is, in fact, the date given on 

 the title page of the Memorias, but the volume was issued in parts, its 

 publication extending over several years, and the issue of the part con- 

 taining Girardinus and Gambusia could not have been earlier than 

 1855. This is evident, as papers written iu 1854 are printed in the text 

 before it. 



In place either of Gambusia or Girardinus the name Heterandria must 

 therefore be used. It has not yet been restricted to either so far as 

 we know. We therefore propose to restrict it to the type of Girardi- 

 nus, regarding Heterandria formosa, Agassiz as its type. Tliis arrange- 

 ment is in accordance with the wishes of Professor Poey, to whom we 



