of the Fishery Board for Scotland. 



37 



In considering, first, the sum of the recaptures — the 13 of long period 

 and the 27 of short period — I find that, adding them to similar records 

 already obtained, yields a total of 63 fish of long migration and 81 fish of 

 short migration. 



There are, however, a certain number of recaptures made on the coast 

 which cannot be regarded as yielding strictly an indication of the return of 

 the fish to fresh water. I have therefore eliminated all coastal recaptures 

 in order to show actual periods of return to fresh water alone, and have 

 at the same time allocated the fish to their districts of marking. The 

 result is : — 



Long 



Period. 





Short Period. 





Tay . 





24 



Tay - 



11 



Helmsdale - 





20 



Helmsdale - 



2 



Brora - 





8 



Brora - 



7 



Naver 





0 



Naver- - 



14 



Deveron 





0 



Deveron 



4 



Borgie 





0 



Borgie 



1 



Spey - 





2 



Spey - 



1 



Ness - 





3 



Ness - 



0 







57 





40 



Six clean grilse marked and released in the Kyle of Sutherland and 

 recaptured as salmon in the same district the following season might be 

 added to the short period, making the total 46, i.e., regarding them as 

 grilse kelts in the winter after marking. 



Still it is seen that, so far as these records show actual and well-defined 

 return to fresh water from the sea, the majority of instances are in the 

 long period column. 



The remarkable thing is that the Helmsdale does not seem naturally to 

 fall into the same category as the Naver, Brora, or Deveron, as one might 

 expect, but approximates to the large river Tay. A considerable number 

 of Helmsdale fish have been taken on the coast showing the short period, 

 or within the limits of the short period. Whether or not further recap- 

 tures will alter the present balance remains to be seen. I am inclined to 

 continue marking and to watch results before formulating any positive 

 view either as to the actual facts, the causes of different habits in this 

 respect in different localities, or the practical issues which may emerge 

 therefrom. The actual results of last season are in singular contrast to 

 some data previously obtained. They show in the restricted application 

 already followed : — 



Naver, 1 1 short. Tay, 7 long and 2 short. 



Borgie, 1 short. Ness, 3 long and 1 short. 



Deveron, 2 short. Helmsdale, 2 long. 



a total of 18 short migrations and 11 long. Others, for the past season, 

 are coastal recaptures. 



I may now refer to the fish which has been twice recaptured, and which, 

 I may add, is still at large. On 3rd March, 1905, it was first marked on 

 the Helmsdale, a rod-caught kelt of 6 lb., 26 inches long. It apparently 

 did not return to fresh water in 1905, but adopted the long migration, 

 remaining over a winter in the sea. On 28th September, 1906, when the 

 Helmsdale river watchers were netting to procure fish to keep in store for 

 the two hatcheries of the district, the fish was recaptured for the first time. 

 It was by this time 13| lb., and 35| inches long. The fish was in 

 unspawned condition. The original mark had been 1396. This was now 



