PEEFACE. 



De. GiJNTHER obsen'es, — " This volume contains the continuation of 

 the families of Physostomi. At the commencement of my account of 

 this order it was my intention to adopt it with the limits assigned to 

 it by Miiller, and to distinguish the Scombresoces as a distinct order, 

 \iz. Malacoptenjgii phari/nf/0(j)iathi ; however, dimng the progress of 

 a detailed examination of these fishes so many points of affinity with 

 the Cyprinodontes became apparent, that I was reluctantly obliged 

 to deviate still more from Miiller's ordinal division. 



" The Salmonidse and the vast literature on this family offer so 

 many and so great difficulties to the Ichthyologist, that as much 

 patience and time are required for the investigation of a single 

 species as in other fishes for that of a whole family. The ordinary 

 method followed by naturalists in distinguishing and determining 

 species, is here utterly inadequate ; and I do not hesitate to assert 

 that no one, however experienced in the study of other families of 

 fishes, wiU be able to find his way through this labyrinth of varia- 

 tions without long prehminary study, and without a good collection 

 for constant comparison. Sometimes forms are met vsdth so pecu- 

 liarly and so constantly characterized, that no ichthyologist who has 

 seen them will deny them specific rank ; but in numerous other 

 cases one is much tempted to ask whether we have not to deal with 

 a family which, being one of the most recent creation*, is composed 

 of forms not yet specifically differentiated. 



" The small family Umbridxe proves to be one of great interest. 

 Represented by two species, one in Central Eui'ope, and the other in 

 I^Torth America, the close affinity of which is here recognized for the 

 first time, it forms one of the most striking instance's against the 

 theory of the geographical continuity of identical forms. 



» No fossil true Salmo is known at present ; tlie nearest fossil approacliing to 

 it is a Mallofna. 



