NATURAL HISTORY OF THE SALMON. 45 



just descended, and those a year yonnger having not 



generally begun to feed at large and show themselves to 



the angler or investigator. Unless the one-year party 



are prepared to maintain that the young fish attain to 



something like full parr size, and move about freely in 



search of food, within a few weeks of their birth, it would 



be hard for them to account for the fact that parrs are 



seen in the rivers in considerable quantity during the 



summer months, after the brood of the former year have, 



according to the one-year theory, gone down to the sea. 



Nevertheless, all this, if forming a strong presumption 



against the one year theory, is not conclusive in favour 



of the two-years' theory ; for the facts are reconcilable 



also, and even more completely, with what may be called 



the half-and-half or mixed theory, which the weight of 



the evidence derived from experiments goes directly to 



support. 



One or two of the facts yielded by Mr. Shaw's experi- 

 ments, and cited (not by himself) as evidence on this point, 

 seem to us defective — those, for instance, in which parrs 

 transferred from the river to the ponds in July put on 

 a migratory dress the next April, for there were no 

 certain means of knowing what was the age of the fish 

 when transferred. The evidence, which equally satisfied 

 and surprised Mr. Shaw, and which alone is admissible, 

 was drawn from the case of the fish hatched in his own 

 preserve, and kept under his own eye, from their birth 

 till their migration. The fish which came out of the 

 ova at Drumlanrig, in the spring of 1837, did not 

 assume the migratory dress, and seek to depart in May 

 1838, as Mr. Shaw had expected, but did so in May 



