RECENT RESEARCH UPON SALMON 239 



There was much rain all July, and a heavy run of grilse. I 

 was absent that month in Norway, and on my return was 

 informed that a number of dark fish had been taken in the 

 stake-nets in the bay. The fishermen spoke of it as a regular 

 occurrence at that season. They considered these fish as 

 having returned to the sea from the upper waters, and they 

 received a lower price for them than for fresh-run salmon. 



I took immediate steps to secure some of these discoloured 

 fish for examination. Unluckily, I was too late, and only 

 succeeded in obtaining one a male which was sent to the 

 Edinburgh Research Laboratory on August 25th, where 

 analysis showed that the skin had begun to turn red and the 

 muscle was pale ; there were parasites in the gills and no sea- 

 lice on the body indications that the fish had passed some 

 time in the river. 



The following year (1901) was an abnormally dry summer, 

 and there was no flood from the first week in June until the 

 very end of August ; nevertheless, early in August, I obtained 

 two or three of these fish, and forwarded them to the 

 laboratory. I have not yet received a report upon their 

 analysis. 



Assuming, as I have no hesitation in doing after hearing the 

 evidence of the net-fishers, that the appearance of these fish in 

 salt water is a regular phenomenon after July floods, it is perfectly 

 clear that their descent from fresh water must be a voluntary 

 migration. A similar seaward movement has been noted as 

 regularly taking place in the Helmsdale ; but that is a rapid 

 Highland river, entering the sea with scarcely any tidal estuary, 

 and the theory might be advanced that salmon, weakened by a 

 long fast, might be washed down out of the upper reaches by 

 heavy summer floods. But in the Cree no such explanation 

 will hold good. In the first place, although the upper reaches 

 of that river are rapid and Highland in character, between 

 them and the sea intervenes the Loch of Cree, fully three miles 

 in length, and above the nets where these discoloured fish were 



