( 496 ) 
scale is figured in fig. 18 and 18a. There can be no doubt but that 
this fish spent two winters in the sea before being caught. On the 
other hand I am by no means so sure with regard to the earlier 
data of this fish. The scale of the grilse of fig. 17, roundabout the 
oldest part, i. e. the one formed during its stay in freshwater, shows 
only a few lines of growth, developed in the first weeks or months 
of the second year, which this fish, before it went to the sea in May, 
still spent in the river and to these rings the parts formed in the 
sea closely joined. In the scale of fig. 18 the part roundabout the 
first year’s scale, which part probably is also formed in freshwater, 
is much broader — but, as is the case in the scale of fig. 17, it 
goes over without any sharp limit into the marine growth of the scale. 
If I had not such good reasons for believing that on the Rhine the 
migration of the smolts to the sea generally takes place in May, and 
if what I have seen of migration in other months (e. g. in autumn) 
was not so extremely scanty, I should be inclined to suggest, that 
this salmon spent a second summer in freshwater, went to the sea in 
autumn, and thence returned to the river, after having passed two 
winters in the sea. This grilse was 3 years and about 5 months old 
on being caught in August, and of that period 2 1 / 4 , or, if the second 
supposition is correct, V/ 4 years were spent in the sea. 
One need not wonder that such cases of uncertainty as the one 
mentioned above still occur. The method of using the scales to determine 
the age of the salmon is comparatively new. Yet I think that my com¬ 
munication, being based partly on observations of the fish and partly 
on the study of its scales, may also contribute to give that method a 
sound basis. Therefore, though my 7 researches are far from finished, I 
thought it useful not to delay the publication of these still provisional 
results. This question with regard to the salmon is a complicated 
one: we have to reckon with the possibility of the parrs staying one 
or more years in freshwater, before they migrate to the sea; with 
the exodus in May or perhaps in another month of the year; with 
their return from the sea in different seasons, and in a very varying 
stage of sexual development. For the' older and larger fishes moreover 
a difficulty arises from the circumstance, that in any case some 
salmon (and perhaps not so few) after having spawned, swim down 
the river again, later to ascend it for the second time. These fish, 
which are called “kelts”, seem to return in greater quantities m 
April and May ; in Holland at least a certain number of them, on 
their way to the sea, are regularly caught about that time : in t is 
year over 100 of these fishes were taken. Now, when these fish wi^ 
have reached their feeding grounds in the sea again, 6 or 9 or 
