( 486 ) 
as the still larger ones of a later year, belong together, which means 
that they date, if not without exceptions, to a large extent from 
the same year of birth. Quite in accordance herewith we see fairly 
regularly, that a year with richer catches of grilse is succeeded by 
a year with more middle-sized salmon, the following year showing, 
not quite so regularly but in many instances distinctly enough, that 
it is still under the same favourable influence. Though exceptions - 
cases which do not enter so well into the scheme — occur, there 
is no longer any doubt as to this rule holding good in general. 
Hence a richer catch of grilse in a certain year has a prophetic 
significance for following years : it opens a period of years with 
rising figures of the catches. Rich years, therefore, are not isolated 
occurrences but form groups alternating with others of poorer years. 
So we understand that a good grilse-year is considered by the 
fishermen also as a good omen for the future.' Hence the study of the 
whereabouts, of the scarceness or abundance in certain years, of these 
smallest of our salmon has always had a great attraction for me. 
Questions like the following: why some salmon enter the river as 
grilse and others not, what the age of a grilse is when leaving the 
sea and entering one of our estuaries, are to be answered before it 
will be possible to have an exact notion of the circumstances which 
are of influence, of the conditions which determine the production 
of grilse and so of salmon in general. This shows sufficiently, I think, 
the great importance of determining the age of these salmon; that 
determination, however, is singularly complicated by the very special 
way of living of the salmon, its being born in the upper parts of 
the river, its passing its youth there, its descending to the sea and 
returning at different ages — a chapter of natural history which 1 
accept to be generally understood by this time, to which, nevertheless, 
these studies of the age based on the construction of the scale may 
add a page of some importance. 
A few words on the determination of that age in general may 
precede. To that purpose the length of the fishes is often made use 
of and this has been done and apparently with some success for the 
salmon also. However, difficulties arose: a very unequal growth was 
found in young salmon reared in hatcheries under evidently quite 
corresponding circumstances: fishes exactly the same number o 
months old may be 5, 7 or 9 cm. long, and young salmon wbic 
have grown up in a free condition may show similar differences m 
size; of course it is very difficult to reckon with such differences 
when the age is to be derived from the size. A second difficulty is 
