ADAPTATION 195 



to point out instances of both in the Characins. 

 I do this fully aware of the fact that some of our 

 experimentahsts have claimed that evidence in 

 favor of mutation would not be noted by the sys- 

 tematic zoologist. It is, however, quite certain 

 that evidence for mutation can not be obtained 

 by experiment only. I have several times found 

 evidence in favor of it in the Characins. 



In the Tetragonopterinae there are parallel 

 genera or subgenera, as we care to look at them. 

 One series has a complete lateral hne; the other 

 series has pores developed on but a few scales. 

 No doubt one has been derived from the other — 

 not once but several times. One species, Hemi- 

 grammus inconstans, is evidently mutating. 

 Two of the four specimens known have a com- 

 plete lateral hne, in the others it is quite short. 



Among hundreds of specimens of another spe- 

 cies with an incomplete lateral hne a single mu- 

 tation has been found with a complete hne. 

 Moenkhausia australe by mutation is producing, 

 or has produced Hemigrammus. In such cases 

 we have, if a bull is permitted, individuals that 

 are specifically ahke but generically different. 



While we have many undoubted cases of muta- 

 tion, there are many reasons why we should not 

 jump to the conclusion that all adaptations have 

 so arisen. 



One example of continuous variation leading 

 to an adaptive point is found in some localities 

 of Nicaragua. Here the species of Astyanaoc 



