70 The Salmon 
time a fish comes into fresh water, it is for the purpose of spawning ; 
so that if a fish comes from the sea three times, it spawns three times. 
The great majority, however, only spawn once. On examining all the 
scales of the marked grilse and spring fish, I find that the contracting 
of the rings generally takes place in January or February, and 
occasionally in March. In April, again, they widen out, and the width 
continues to become greater during the summer and autumn months. 
By referring to some of the photographs the reader will observe 
when a fish has spawned as a grilse, gone down as a kelt, and come 
up as a clean salmon. Count the number of rings that have been put 
on from the time it was marked to the time it returned, and refer to 
the number and weight of the fish (Fig. 65). Due allowance must, of 
course, be made for kelts lingering in the river after being marked. 
The time need not be extended beyond the 1st of May, as most kelts 
have disappeared by that time. In almost every instance you will tell, 
within a few weeks, the time the fish has been in the sea. 
I will now deal with the scales of small spring fish which are of 
the same age as the grilse of the previous year, and will confine my 
remarks to fish marked by us, including the autumn run as well. 
The first one caught was on 18th February 1907, and its weight 
was 9» lbs. (Fig. 66). Now if this grilse, caught on rst June 1906, 
had fifty-one rings on its scales, how many should this February fish 
have? The answer is sixty-two. Now at the same date the kelt of 
the grilse of the same age has only fifty-one. This kelt grilse, being 
on its way to the sea, will add another sixteen rings if it remain in 
the sea for a year; so that when the spring fish of 18th February is 
returning to the sea as a kelt, the kelt grilse will be going up the river 
as a clean fish, and will now have sixty-seven rings on its scale, five 
more than the 18th February spring fish kelt which is going down. 
The last marked fish we got of the same age was caught on roth 
September 1907, and the number of rings on its scale was seventy-two 
(Fig. 77). From the enlarged photographs of the scales taken each 
month, it will be seen, in addition to the number of rings put on each 
month, where the lines are contracted in some cases, and also where 
