60 ATHERINID^ : SILVERSIDES 



91. Thyrina sardina (Meek), 1907.-^*- 



Chirostoma guatemalensis Gill and Bransford, Proc. Acad. Nat. Set. Phila., 

 187, 1877. 

 Thyrina guatemalensis Fowler, ibid., 40, 736, pi. 43, 1904, lower figure (ap- 

 parently not Atherinichthys guatemalensis Giinther). 

 Melaniris sardina Meek, Publ. Field Miis., (Zool.), 7, 114, 1907. 

 Thyrina sardina Regan, Biol. Centr.-Amer., Pisces, 189, 1908. 

 Type-locality. — Lake Managua, Nicaragua. 

 Range. — Lakes Nicaragua and Managua, Nicaragua. 



Thyrina sardina has the mouth more obUque than the other species 

 of the genus, and the rami of the lower jaw are much more elevated 

 within the mouth, forming an apical angle which is scarcely obtuse. 



If it should be found necessary to merge Thyrina with Menidia, 

 then this species would require a new name, there being a prior Menidia 

 sardina. 



92. Thyrina guatemalensis (Giinther), 1864. 



Atherinichthys guatemalensis Giinther, Proc. Zool. Soc. London, 151, 1864; 

 Trans. Zool. Soc. London, 6, 443, 1869. 

 Menidia guatemalensis Jordan and Evermann, Bull. U. S. Nat. Mi4s., 47, pt. 1. 



801, 1896. 

 Thyrina guatemalensis Jordan and Evermann, /. c, pt. 3, 2840, 1898; Regan, 

 Biol. Centr.-Amer., Pisces, 64, 1907 (in part) ; ? Ann. Mag. Nat. Hist (8), 12, 

 472, 1913 (Rio Condoto, Pacific slope of Colombia). 

 Type-locality. — Lakes of Huamuchal, Pacific slope of Guatemala. 

 Range. — Pacific slope of Guatemala, ? southward to Colombia. 



93. Thyrina balsana (Meek), 1902.-^*- 

 (Plate IV, Fig. 17) 

 Melaniris balsanus Meek, Publ. Field Mus. (Zool.), 3, 117, pi. 28, 1902; 5, 183, 

 fig. 64, 1904; Jordan, Guide to the Study of Fishes, 2, 218, 1905. 

 (Plate V, Fig. 17) 

 Type-locality. — Rio Balsas, at Balsas, Guerrero, Mexico. 

 Thyrina guatemalensis Regan, Biol. Centr.-Amer., Pisces, 64, 1907 (in part). 

 Range. — Streams of the Balsas Basin, Mexico. 



This species and sardina, which Dr. Meek referred to his genus 

 Melaniris, agree with the type of Thyrina (evermanni) in all es- 

 sential respects. In both of these species the belly is apparently com- 

 pressed as though pinched into that form, but the abdomen is distended 

 with mature eggs in the specimens at hand. 



T. BALSANA is vcry closely related to guatemalensis, evermanni, 

 and crystallina. We have no specimens of guatemalensis at hand, 

 and GiJnther's description is incomplete. The only point of difference ap- 

 parent lies in the size of the scales : in ten paratypes of balsana there 

 are 38 to 40 — 9 or 10 rows, while Giinther counted 36-7 in guatema- 

 lensis. T. balsana differs from evermanni in the less sharply com- 



