379 



The low percentage of impregnation, on the other hand, must be re- 

 gavded in most cases, I feel sure, as due to ant'avorable conditions of the 

 milt, and in some cases to the unripe condition of the eggs. Males that 

 have passed the height of their breeding season, or which may have been 

 less able to endure the conditions of confinement in aquaria usually show a 

 reduced fertilizing ])ower compared to 'perfectly fresh and ripe individual. 

 The testes were in all the experiments cut out, so that it is quite probable 

 that in many cases imperfect milt was used. I was, furthermore, not able to 

 establish any constant difference in the percentage of impregnations in re- 

 ciprocals. Allowing for the influence of the condition of the milt in deter- 

 mining the percentage of impregnation, in all cases where a fair trial was 

 made in reciprocal crossing of two species it was approximately as high in 

 one direction as the other. It is interesting to note here that Kammerer 

 07 using fresh water flshes, found, among the few forms he used, two crosses, 

 Perea flnviatilis x Acerina schi-aetser and Lucioperca stindra x Perca fluvia- 

 tilis, in which it was possible to impregnate when the first named in each 

 case was the male, but not if female. It is also impossible to fertilize the 

 eggs of Aspro zingel with the milt from the following nearly related forms : 

 Perca flnviatilis, Lucioperca sandra and Acerina sp?, but was able to fer- 

 tilize them with the milt from the distantly related form Cottus gobio. It 

 would seem from these experiments that fresh water fishes lend themselves 

 less generally to hybridization than the marine species. 



Kammerer's statement that the eggs of Aspro zingel are fertile to the 

 sperm of the distantly related form Cottus gobio when they were immune 

 to the three nearly related forms above indicated, because Cottus had a 

 similar habitat, and had with this also acquired the power to fertilize this 

 species is. of course, a mere fancy. If he had tried to cross this form with 

 other distantly related forms he would probably have found that they, too, 

 would fertilize the eggs regardless of their habitat relationship. 



DEVELOPMENT. 



In my study of the development of these various hybrids I have not 

 attempted to get a complete morphological picture, nor have I paid much 

 attention to the inheritance aspect. I have regarded development rather 

 from a physiological standpoint. The main points of interest, therefore, 

 have been, first, how generally and within what limits can the sex-products 

 of the various forms of teleosts be grafted upon each other, so to speak, 



