ATLANTIC SALMON 



21 



and the particular river from which the fish 

 emerge. 



Great difficulty exists when one attempts 

 to estimate the speed at which salmon swim 

 in the sea. A direct comparison of mini- 

 mum length of journey and the time interval 

 between marking and recapture can give 

 only the slowest possible average speed. We 

 can not know the exact course followed by 

 the fish ; nor can we estimate the time which 

 may be spent after marking before the 

 journey is started or before final capture 

 after completion of the main passage. Re- 

 sults do suggest, however, that when the fish 

 is engaged in a long journey it makes the 

 passage in more direct fashion than when a 

 shorter trip is in prospect. The average 

 speeds obtained for the longer trips are al- 

 most always higher than those for journeys 



of only a few miles, and the error introduced 

 by calculations based on days absence is also 

 smaller for the longer time intervals. For 

 instance, a fish marked one morning and re- 

 captured the following afternoon would be 

 counted as absent one day, whereas it has 

 taken one and a half days for the journey — 

 an addition of 50 per cent to the nominal 

 time. Whereas if the total time be say 12 

 days the error of a half day becomes of 

 much less importance. On the Scottish 

 coast salmon may accomplish journeys 

 greater than 100 miles at an apparent speed 

 of between 20 and 35 miles per day ; the lat- 

 ter rate occurs only in the case of two fish 

 both of which travelled more than 250 miles. 

 In general, Professor Dahl has found the 

 same rates of speed although he has a 

 greater number at the maximum rates. He 



Fig. 8 



