38 



BULLETIN OF THE BTJBEATJ OF FISHERIES 



accordingly, the proportions in which each age group is present among the progeny of 

 a certain spawning. By comparison of similar figures from successive spawnings, we 

 shall be prepared to develop the amount of variation that is found in this relation. 

 If, for instance, among the progeny of the 1921 spawning we find that the 6-year age 

 group comprises 20 per cent of the whole, how will this compare with the 6-year 

 group that develops from the 1922 spawning, and this with the results of successive 

 years? Will they also be present in approximately 20 per cent of the whole, or will 

 the relation be a widely varying one in successive broods? If in the relative numbers 

 of each age group a fair degree of uniformity is found to develop from different 

 lots of eggs, the number of individuals found in the younger age groups of any brood, 

 which are the first to make their appearance in the runs, will serve as a reliable basis 

 for prophecy concerning the size of successive age groups that develop from this 

 same lot of eggs. From the size of these early samples of the brood we should be 

 able to compute with a fair approach to accuracy the sum total of all the progeny 

 of this spawning and would find ourselves in possession of that great desideratum — a 

 reliable basis for prophecy concerning the size of future runs. If, however, the rela- 

 tive number in various broods that develop at the same age is highly variable, the 

 size of the younger age groups, as they successively appear, would present but 

 indifferent evidence concerning the size of the older age groups, some of which may 

 well constitute the dominant group or groups in future runs. If such wide variation 

 exists, we can only seek for causative factors and may succeed in correlating a 

 tendency in a certain brood toward early or late maturing with a certain set of external 

 conditions. 



For convenience of reference in connection with the following discussion of 

 dominant and other age groups and the brood years to which they are referred, we 

 present below the records of Karluk red-salmon packs, in even thousands of cases 

 (48 one-pound cans to the case), from 1910 to 1920. We have no evidence con- 

 cerning the size of the spawning escapements for these years. 



We present below our analyses of the runs of the various years for which we 

 have data, taking them in chronological order. 



Cases Cases 



1910 107,000 



1911 124, 000 



1912 89, 000 



1913 62, 000 



1914 39, 000 



1915 59, 000 



1916 167, 000 



1917 166, 000 



1918 78,000 



1919 78, 000 



1920 98, 000 



RUNS OF 1916 AND 1917 



For a period of four years immediately preceding 1916 the Karluk pack of red 

 salmon had been very poor, ranging from 39,000 cases in 1914 to 89,000 cases in 

 1912. In 1916 and 1917, however, the pack suddenly increased to over 165,000 

 cases in each of those years. It becomes of interest to ascertain the prevailing 

 age of the 1916 and 1917 fish and to discover which brood years had been mainly 

 responsible for the sudden increase in the runs of those two years. 



Our samples of the 1916 fish are but two in number — one of 116 fish taken on 

 July 19, the other of 266 fish obtained August 24, An analysis of these is given in 



