Salmonidae of Norway. 79 



(3) That salmon which have spawned in situations 



from which departure is impossible until the 

 following spring, in a river which has no autumn 

 run, spawn but once in two years. 



(4) That fish which ascend the rivers each autumn 



can well spawn each year. 



(5) That fish can spawn each year in rivers which 



admit of both spring and autumn " runs," pro- 

 vided that the fish do not confine themselves 

 to either spring or autumn " running." 



(6) That fish which descend the rivers as kelts in 



winter (December and January) can reascend 

 as fresh fish in the following summer (July and 

 later). 



I call attention to these points, as, if I am correct, 

 they bear somewhat strongly upon the system of breed- 

 ing from " spring " or " autumn " fish, and upon the 

 annual growth of the salmon. 



The results derived from the marking of fish by Mr. 

 Walter E. Archer, than whom there exists no greater 

 authority upon the natural history of Norwegian salmon, 

 are most instructive, both as regards the facts they prove 

 and the inferences to be drawn from them. 



In the Suldal (or Sand's river), the Aaensira, the 

 Figgen, and the Topdals rivers, of some thousands of 

 fish marked, 56 were recaptured. 



