of the Fifhery Board for Scotland. 



15 



fish, therefore, which were present at that time in the upper waters 

 must have entered the river, at any rate, before tlie end of August. It 

 also seems to lay the foundation for the complete solution of the 

 question of whether the early fish form the breeding stock in the upper 

 waters, leaving the lower reaches for the later arrivals. For, if it is the 

 these investigations seem to show, (1) that there is a marked 

 difference in the size of fish entering the rivers in different periods of 

 the year (Special Report, pp. 10 and 69), and (2) that they do not 

 increase in length during their sojourn in fresh water (Special Report, 

 p. 171), then, given the average length of fish entering the river in 

 each period, and that of fish frequenting the different tributaries oi- 

 parts of the main river at spawning time, it will be possible to show 

 whether or not fish frequenting any tributary, or part of the main river, 

 correspond in length with those entering the liver in any particular 

 season of the year. 



Dr. Noel Paton refers in the second place to the very marked difference 

 in the weight of the muscle of fish in the upper waters in October and 

 November, as compai-ed with those coming in from the sea, as an 

 indication that the immigration of the more muscular fish from the sea 

 to the upper water practically ceases in the autumn (Special Report, 

 p. 75). But, as was pointed out when discussing the length of fish 

 taken in the upper w^aters, the fact of fish with greater muscular 

 development not being sent from the upper waters for examination does 

 not prove that they were not there, seeing that the fish sent were 

 selected of a particular size, and, as has been shown, the more muscular 

 fish coming in from the sea in the autumn were larger than was 

 required, and would probably, therefore, not have been sent so long as 

 the smaller-sized fish were procurable. 



The marked difference in the muscular development of the upper and 

 lower water fish respectively does, however, appear to tell strongly against 

 the theory of any general to-and-fro migration between river and sea. If 

 it is the case, as these investigations seem to show, that salmon continue 

 to feed in the sea, and cease feeding, or, at any rate, in utilising any 

 food which they swallow, after they enter the rivers (Special Report, 

 p. 171), then it follows that the longer they remain in fresh water the 

 greater will be the loss of muscle. If, therefore, any such to-and-fro 

 migration were general, we should expect that the amount of muscle 

 expended in reaching the upper waters would in each period have been 

 about the same, and not, as is shown in Table V^II. (Special Report, 

 p. 67), that the differences in the weight of the muscle between fish 

 taken at the mouth and those taken in the upper w^aters should become 

 more and more marked as the season advances. The ever-increasing 

 difference as the season advances in the weight, weight of muscle, and 

 weight of solids in the muscle would seem to indicate that, even if seme 

 of the more muscular fish coming in from the sea in the autumn ascend 

 to the upper waters, and smie of the salmon which enter ed the rivers 

 early in the year return again to the sea, among those frequenting 

 the upper waters from May to November there are a large number 

 which lemain there drawing upon the material stored in their bodies 

 for the construction of their ovaries and for their supply of energy 

 (Special Report, pp. 66, 77, and 86). 



Lastly, the marked difference w^hich is shown in the weight of the 

 ovaries in upper-water as compared with lower-water fish seems to show 

 that the same fish are not passing to and fro between river and sea, but 

 rather that the fish supplied from the upper waters in October and 

 November with the large weight of genitalia and the small muscular 



