46 BULLETIN OF THE NATURAL HISTORY SOCIETY. 



from snout to caudal, forming a broad conspicuous patch 

 on operculum. Head three and two-thirds, and depth 

 four times in body. Sides below lateral band rosy. 

 Females larger and more richly coloured than the males. 

 Gill-membranes not forming a fold at the isthmus. 

 Branchial leaflets stouter, oblong not acute, filling gill 

 cavity completely and causing it to bulge. Gill rakers 

 shorter, less acute and with broader bases. Teeth 2.4 — 

 4.1 or 4.0, more acute and hooked. The Garnett lake 

 variety are almost the counterpart of the Maugerville 

 specimens, reaching a length of five inches. 



While these differences may not be entitled to specific 

 recognition, it must be remembered they are constant, and 

 mnny species have been established upon fewer and less 

 important characters ; nevertheless, the writer believes 

 that this principle has been pushed to an unwarranted 

 length, needlessly multiplying species and perplexing the 

 "whole science of ichth^'ology. Just how far modifi- 

 cation must extend, and what structural and other 

 diversities must exist before a species can be declared, 

 will always be a subject of contention ; still it behooves 

 men of science to be as conservative as possible, and slow 

 to stamp with natural isolation and finality forms which 

 are but initial steps in that process of development which 

 matures in the production of well-marked species. These 

 little fishes furnish a good illustration. They occur only in 

 small ponds, widely separated, and from which predaceous 

 fishes are entirely excluded, each set having a very re- 

 stricted range with conditions of life peculiar to itself. 

 As the creature is but the vitalized history of its environ- 

 ment, when the latter diflers in two or more cases, dis- 

 similar products must result. In this way what may 

 have been in past ages a fairly uniform species, scattered 

 over an immense area more or less accessible to all, has, 



