100 Annals of the Carnegie Museum. 



These accounts are In almost entire accord and are the most com- 

 plete yet given. The manner of conducting the spermatozoa from 

 the genital opening to the tip of the anal, however, is yet to be made 

 known. In most of the viviparous forms, except Fitzroyia and 

 Anableps, the genital duct of the male opens immediately in front of 

 the base of the anal fin. In certain genera, such as Pcecilia, Molli- 

 enisia, Limia, and Xiphophorus, the ventral fins, which adjoin the anal, 

 are elongated, and, if pressed against it, would form a tube extending 

 nearly to the end of the anal, through which the spermatozoa might 

 pass. In the majority of the viviparous genera, however, such as 

 Heterandria, Phalloceros, and Cnesterodon, the ventral fins are alto- 

 gether too small to be of such service. Further observations on living 

 fishes will be necessary to determine the manner in which the sperm 

 bodies are transferred to the tip of the anal. In Fitzroyia and Ana- 

 bleps the ureter and the sperm-duct continue to the end of the anal, 

 which in this case is an actual tube. 



Garman (c, p. 1012) observed that in Anahleps the tip of the tubular 

 anal of the male is invariably pointed either to the right or to the 

 left. In females the genital orifice is protected by a large scale, which 

 is fastened either on one side or the other, and permits the entrance of the 

 anal of the male only from the opposite side. This Garman fancifully 

 thought wa? a unique device to insure cross-fertilization. It is obvious 

 that a male with the tip of the anal directed toward the right can mate 

 only with a female having the left side of the genital orifice free through 

 the fastening of the protective scale upon the right side. Garman 

 thought that in the same brood probably all males and all females were 

 of the same type of structure and that interbreeding would thus be 

 prevented. In Fitzroyia the anal of the male is also tube-like and the 

 direction of its tip is fixed. But since in this genus the genital orifice 

 of the female is unprotected by a scale or otherwise, it seems probable 

 that mating can be accomplished by either type of male with any 

 female. There would thus be no provision to insure cross-fertilization 

 in this genus. 



Philippi observed that, although both sides of the anal are alike 

 in P. caudomaculatus and C. decem-maciilatiis, the anal can in any one 

 individual be used only on one side of the body. In P. caudomaculatus 

 it is prevailingly the left side, and in C. decem-maculatus the right, from 

 which the anal may be thrust out. This it is presumed means that 

 males of C. dcceni-maculatiis approach the female from the left side, 



