334 H - H - NEWMAN. 



gradually fading out (Text Plate V., Fig. 3). The lower line of 

 processes behaves in a slightly different manner. Instead of 

 spreading out to form a continuous straight band it forms a series 

 of short, nearly longitudinal bars (Text Plate V., Fig. 4). Later 

 on these short bars fuse end to end in various fashions to form 

 from two to five longer bands or stripes one of these extending 

 nearly two thirds of the length of the body and the others, some 

 above and some below the lower long stripe, vary greatly in length 

 in different individuals (Text Plate V., Fig. 5). This process of 

 conversion of bars into stripes takes place, as in the case of the 

 doubling of bars in the male, in an anteroposterior direction. In 

 half grown females all stages of the process may readily be found. 

 Such specimens show from two to five of the posterior bars intact 

 while the stripes at the anterior end are well marked. I have never 

 yet seen an adult female in which all of the bars had disappeared, 

 the most posterior one always persisting. The appearance of 

 adult females is strikingly characteristic and must serve as a very 

 efficient recognition mark for the males. The sexual difference 

 is accentuated by the fact that the fins of the female are nearly or 

 quite devoid of any dark pigment, only a small amount of a light 

 yellowish pigment being present and that chiefly on the caudal fin. 



The fins of the male bear the same characteristics as do those 

 of F. heteroclitus, but are even more pronounced in their sexual 

 dimorphism. Especially is this the case in the anal fin which is pro- 

 longed backward more markedly than in the previously described 

 species (see Plate XXVII., Fig. 4). The anal fin of the female 

 is swollen and inflamed in egg-laden specimens freshly brought 

 in. The flow of milt is free and copious only in males possessing 

 the highest coloration. The contact organs are even better de- 

 veloped than in F. heteroclitus and their distribution is very similar. 



Beyond these indications I have no direct evidence bearing on 

 the spawning behavior of Fitnduhts inajalis. Although taken in 

 large numbers when they were apparently at the height of their 

 sexual activity, as indicated by the abundance of ripe eggs and 

 sperm, they showed no tendency to spawn in captivity. The 

 fish are much wilder than F. lieteroclitns and seem to feel cap- 

 tivity much more keenly. All of their normal activities seem to 

 be, for a time at least, inhibited by confinement in restricted 



