638 FISHES. 



found very constant (see Fig. 56, p. 131). If unexpected 

 variations occur, their cause may be found in a partial con- 

 fluence of the caeca, as we have observed that specimens of 

 S. Uv&nensis (a species normally with from 70 to 90 cseca), had 

 those appendages of unusual width when the normal number 

 was diminished. 



We have mentioned above that many points in the life- 

 history of the Salmonoids still remaia very obscure : — 



1. Johnson, a correspondent of Willughby (" Hist. Pise," 

 p. 194), had already expressed his. belief that the different 

 Salmonoids interbreed ; and this view has since been shared 

 by many who have observed these fishes in nature. Hybrids 

 between the Sewin {S. cambricus) and the Eiver Trout (S./ario) 

 were numerous in the Ehymney and other rivers of South 

 Wales, before Salmonoids were almost exterminated by the 

 pollutions allowed to pass into those streams, and so variable 

 in their characters that the passage from one species to the 

 other could be demonstrated in an almost unbroken series, 

 which might induce some naturalists to regard both species 

 as identical. Abundant evidence of a similar character has 

 accumulated, showing the frequent occurrence of hybrids be- 

 tween S. fario and S. trutta ; hybrids between S. fario and 

 species of Charr have been abundantly bred by continental 

 pisciculturists. In some rivers the conditions appear to be 

 more favourable to hybridism than in others, in which 

 hybrids are of comparatively rare occurrence. Hybrids be- 

 tween the Salmon and some other species are very scarce 

 everywhere. The hybrids are sexually as much developed 

 as the pure breed, but nothing whatever is known of their 

 further propagation and progeny. 



2. Siebold has shown that some individuals of every 

 species are not sexually developed, and that such individuals 

 differ also externally from those normally developed. How- 



