48 ATHERINID^: SILVERSIDES 



lateral band is no wider than in I. rhothophilus). These marked dif- 

 ferences being- less striking than the numerous points of resemblance 

 between them, we retain these two remarkable atherines in the same 

 genus. ^_ ;, - i ^^\ ,M\. 



'i/vO|M;^^79^jso natalensis Regan. 



Iso NATALENSIS Regan, Ann. Durban Mus., 2, 200, 1919. (Durban, Natal.) 



Subfamily ATHERINOPSIN^ Fowler. 



AtheriniN/E Fowler, Proc. Acad. Nat. Sci. Phila., 40, 727, 1904 (in part). 

 Atherinopsin^ Fowler, /. c, 737. 



This large subfamily is restricted to the two shores of America, and 

 includes all of the American species of the family, excepting a few of 

 Atherina and Hepsetia. Several of the genera are largely or wholly 

 composed of fresh-water species. 



Excepting the lack of vomerine teeth, the genus Menidia, especially 

 the subgenus Ischnomembras, possesses all of the characters which one 

 would postulate the ancestral type of this group to have. This point is 

 emphasized in the analytical key to the genera. From some Menidia- 

 like form several distinct lines of evolution may be traced to aberrant 

 genera. Thyrina, Archomenidia, and Chirostoma, Mexican fresh- 

 water genera, are derivable from such a fish. Thyrinops, considerably 

 resembling Thyrina, leads to the aberrant Atherinella; the remark- 

 able genus Eurystole, convergent toward the unrelated genus Iso of 

 the Atherinin^, also shows points of resemblance to Thyrina. 

 HuBBESiA and Membras, and also Leuresthes and even Labidesthes, 

 are other modifications of the Menidia type. From Menidia likewise, 

 the large species of temperate South America, belonging to Austro- 

 menidia and related genera, have doubtless been derived. The highly 

 peculiar fresh-water Xen atherina was apparently also derived from 

 a fish like Menidia or Thyrina, but the intermediate stages are un- 

 known, and probably extinct. The other genera of the group agree in 

 having the premaxillaries non-protractile; they form the subfamily 

 Atherinopsinte as defined by Fowler. We are not certain, however, that 

 the North American genera of this type, namely Atherinopsis and 

 Atherinops, have had acommon origin with those of South America 

 (Basilichthys). It is fully evident, however, that all belong to the 

 Menidia group of genera. For this entire group we use the only sub- 

 family name that has been proposed — Atherinopsin/E. 



We are unable to place generically two American species, the 

 descriptions being incomplete and probably inaccurate in detail. Athe- 

 rina microps Poey ^^ from Cuba was accepted as a true Atherina by 



15 Mem. Hist. Nat. Cuba, 2, 266. 1861. 



