386 



culation is a vital stage in the proper progressive development of the 

 embryo and is followed in normal embryos by a period of rapid growth, the 

 question still remains, why does the circulatory system fail to develop 

 properly? Why do we have so. many embryos stop their development before 

 the period of heart formation, and why do we have so commonly failures 

 to emerge from egg, or die shortly after, when in the latter the circulation, 

 at least to all appearances, has been normal? If we consider the experi- 

 ments tabulated in Table 9 from the view point of the correlation between 

 the degree of development and the relationship of the species combined, 

 we see at once that only those species that belong to the same genus, or to 

 very closely related genera, will produce hybrids that develop to the point 

 of hatching. Even within this group a difference in this respect can be 

 observed between species very closely related, and species more distantly 

 related. Thus Fundulus heteroclitus combined in either direction with 

 Fundulus diaphauus will produce a large proportion of free swimming em- 

 bryos. These two species, although the former is a marine and the latter 

 fresh water, are structurally very closely allied. Fundulus majalis is much 

 less closely related to Fundulus heteroclitus, although belonging to the 

 same genus. When the latter is taken as the female a large proportion of 

 vigorous fry are obtained. The reciprocal has never yielded me embryos 

 that would emerge from the egg, although, with the exception of the yolk 

 bag, normal in appearance. Then the species used belong to separate 

 genera the proportion of embryos that emerge normally is, as a rule, much 

 smaller than in the preceding condition. 



All species that are more removed from each other than closely related 

 genera, fail to produce hybrid embryos that will complete development to 

 the point of hatching. Among this latter group of hybrids the stage to 

 which development is carried varies considerably in the different com- 

 binations. This, too, can be roughly correlated with the relationship of the 

 species combined, so that two species belonging to distantly related orders 

 like Fundulus lieternclitus x Tautogolabrus adspersus give rise to hybrids 

 that can not go much beyond the closure of the blastopore, while if the same 

 form is crossed with its nearer relative. Menidia notata, development pro- 

 ceeds very much further although stops I'ar from the point of hatching. 

 This will be further taken up below. 



We produce, then, among fishes a series of hybrids that range in 

 success froju those in which none of the embryos develop very much be,- 



