PARR AND SMOLTS in 



before it descended as a smolt. The two scales are shown side by side 

 for convenience of reference. 



Now it has to be admitted that the British data are too limited in 

 extent for any certain principles to be deduced from them, but I think 

 they prove conclusively that there is considerable variation of the time 

 of descent of the sea-trout smolts even in any one particular district or 

 river. They also suggest, if they do not actually prove, that the period 

 of descent of the majority of the smolts, in Scotland and England as 

 well as in Norway, is after three winters' residence in fresh water. 

 One Loch Lomond instance also proves that a sea-trout may voluntarily 

 delay its descent till after its fifth winter spent in fresh water, while a 

 single instance given by Mr. Dahl also proves that in Norway the delay 

 may be protracted even for another year. There can be no question, 

 however, that the general principle is established from these data that 

 one has no right to assume as regards any river that all the sea-trout 

 smolts which descend to the sea in any particular year are the produce 

 of any one particular preceding spawning season. 



It is an interesting speculation — and I think a legitimate specula- 

 tion — that the occurrence of a lake in any river system may encourage 

 the young fish to protract their stay in fresh water. The data concern- 

 ing Loch Lomond fish are at least not adverse to that assumption, and 

 I am informed by Mr. Hutton that two considerable lakes interrupt the 

 uniform channel of the Osen river in Norway. A wider investigation 

 may establish this theory which on the face of it does not seem to me 

 to be unreasonable. I shall return to this point later.* 



Some connection with what I have called Mr. Regan's theory — 

 though of course it is not alone his — may, to speculate further on these 

 data, be traced in this way. 



" If salmon smolts," writes Mr. Calderwood, " are retained in fresh 

 water beyond their natural time of descent to the sea, they assume again 

 the trout-like appearance of the juvenile." This " natural time of 



1. See post, page 144. 



