GROWTH AND MATURITY OF SALMON IN THE OCEAN 
47 
fish vsdth the stream type of nuclear growth and 65.4 per cent with the ocean type. 
In 1917, however, the same author (Fraser, 1921) found exactly the same per- 
centage of fish with the ocean type of nucleus which we have recorded here — 78 
per cent. 
Contrasted with this condition of relative stability in the percentages of fish 
with ocean and with stream nuclei, which is found among the fish taken in the 
ocean, are the marked variations found among the fish taken in the river and which 
are illustrated in Figure 12. There would seem to be but one possible explanation 
of these facts, namely, that the fish taken outside the mouth of the river repre- 
sent a fairly accurate sample of the entire population of Columbia River chinooks 
contained in the ocean, and that these average percentages of fish with ocean and 
with stream nuclei represent, approximately, the percentage of fish with stream 
and with ocean nuclei that would go to make up the total of the run in the Colum- 
bia River. It is, of course, possible that races other than those belonging to the 
Columbia River Basin are taken off the mouth of the Columbia River, but since 
the Columbia is by far the most important chinook salmon stream on the coast it 
can hardly be doubted that the great majority of fish found just outside are native 
to this river. The fish found outside therefore represent the supply of fish from 
which the runs of mature fish found in the river are drawn. It would be expected 
that the removal from this supply of considerable numbers of fish with stream 
nuclei during the early part of the run, and of greater numbers of fish with ocean 
nuclei during the later part of the run, would tend to upset these ratios in a sys- 
tematic manner, but this effect is not noticeable. The high percentages of imma- 
ture fish taken outside would tend to obscure this effect, since the migration of 
mature fish would affect only the percentages of fish with ocean and with stream 
nuclei among the maturing fish. It seems probable, however, that with many more 
data than are available, some systematic fluctuations in the proportions of fish 
with ocean and stream nuclei might be shown. 
On account of the fact that the variations in the percentages of fish with 
ocean and with stream nuclei among the fish taken in the ocean are practically 
negligible, the average percentages of fish of the different age groups taken through- 
out the season is significant. (This, as mentioned above, was not true of the data 
from the fish taken inside the river, on account of the great variations in the per- 
centages of fish with ocean and "with stream nuclei and the differences in the rela- 
tive importance of the different parts of the run. See page 44.) These data are 
given in the bottom line of Table 10, and from them it is evident that among the 
fish with ocean nuclei the 3-year group is most abundantly represented, forming 
31.8 per cent of the total take. The 4-year group comes next and forms 23.8 per 
cent of the total. Fish in their second year with ocean nuclei form the next most 
important group, with an average of 16 per cent. A few 5-year fish and a very 
few 6-year fish appear in the collections. In the case of the fish with stream nuclei 
it is seen that 3-year-old fish are most numerous, followed by 4, 5, 6, and 2 year 
fish, in the order of importance. Fraser (1920, p. 172) gives the percentages of 
fish of different ages of "sea type;" that is, those whose scales show the ocean type 
of nuclear growth, as observed in the Straits of Georgia. The figures are very 
» 
