396 



H. H. NEWMAN 



least as wide a cross as anj^ possible within the bounds of the 

 genera given in this table is that between the first and the last 

 genera, Fundulus and Scomber. In general it may be said that 

 any Teleost will cross with any other with the exception of vivi- 

 parous species, and those with peculiar breeding or brooding 

 habits, such as the pipe-fishes, where secondary obstacles to cross 

 insemination present themselves. It is also worthy of note that 

 no artificial aid to insemination is necessary in any cross. 



In the case of crosses between closely allied species, as, for 

 instance the various species of Fundulus (F. heteroclitus, F. ma- 

 jalis, F. diaphanus), in which paternal heredity is as obvious in 

 many ways as is maternal, in that paternal characters are often 



TABLE 2 

 Genera of Teleostei used in hybrid experiments 



dominant over maternal. The same is clearly true for crosses 

 between two genera of the same family as when the genus Gas- 

 terosteus is crossed with Apeltes, both genera belonging to the 

 family Gasterosteidae. In this case a large percentage of healthy 

 hybrid larvae hatch and they are by no means pure maternal. 

 Crosses of interfamily width such as that between representatives 

 of the different families of Acanthopterygii give similar results, in 

 so far as success in development and strength of paternal heredity 

 are concerned, as do crosses of suborder width and need not be 

 considered at length. As a very typical case of heterogenic 

 hybridization of suborder width ' among the Teleosts I have 

 chosen that between Fundulus heteroclitus of the suborder 



