90 Appendices to Twenty -second Annual Report 



March of last year, 1903, it is taken by rod, a salmon of 19 lbs. Similarly, 

 8306 was marked in the Tummel, a female (grilse ?) kelt of 6 lbs. in Jan- 

 uary 1902. In March 1903, it is netted in the lower Tay a clean fish of 

 18 lbs. The spring fish of the Tay are a well-defined class and commonly 

 scale 18-20 lbs. There is also, however, a spring run of small 

 fish. When we regard the increase of weight and the undeveloped 

 condition of the genitalia in these large spring fish, it seems improb- 

 able that in the interval between our marking and recapture the 

 fish have been in fresh water. If this is so these fish are only 

 in their second ascent, and are what are termed "gillings" in the 

 Severn.* By our marking of grilse kelts during the annual close time, 

 we have noticed that these fishes after spawning return to the sea more 

 rapidly than salmon seem to do in a large river such as the Tay t On 

 28th November 1901, for instance, four grilse kelts were taken in the 

 tidal water below Perth, one being as far down as the Bush fishing 

 station opposite the mouth of the Earn. I do not mean to imply by 

 this that grilse spawn earlier than salmon ; unspawned grilse are to be 

 found in fresh waters throughout the spawning season. What one 

 notices is that in netting after the bulk of the fish have spawned, one 

 gets very few grilse kelts. Marked salmon kelts (females) are recaptured 

 after one to two and a half months in fresh water ; marked grilse 

 kelts are not again seen on their descent to the sea. Only one case 

 has, I think, occurred, and that was a grilse both marked and recaptured 

 in the estuary; No. 3177, marked at Tappie Station on 16th January 

 1902, recaptured at Pyeroad Station on 5th February 1902. In support 

 of this result also, we have an unspawned grilse caught, on its ascent of 

 the tidal water, on 16th January 1902, at Flookie Station, recaptured a 

 kelt already in the tideway on 7th February 1902. 



Now, let us examine one other case. No. 9402 was a grilse kelt of 4 

 lbs., marked at Logierait in the upper Tay, on 5th February 1903. 

 This fish was recaptured, clean, in the estuary of the Tay, at Flookie 

 Station, on 31st July 1903. It was then a salmon of 10^ lbs., having 

 increased 6| lbs. in weight and 5 inches in length. 



From this one class of grilse, therefore, it appears that we may have 

 two different seasonal runs of salmon, one in the first summer following, 

 and the other in the next spring. Diagrammatically, the condition might 

 be represented thus : 



Grilse Kelt to Clean Salmon. 



Spring Salmon. 

 fdOS. 



This is a clearly defined example of a divided migration, as set forth 

 in 1860 by Mr. John Dickson, then agent for the Tay Fishery 

 Proprietors, in an article to the Perthshire Courier, and afterwards by 



* Willis Bund, Salmon Problems, p. 97. 



t Twentieth Annual Report of the Fishery Board for Scotland,' Appendices to 

 Part II., p. 96. 



