66 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 94 



Five 23-year cycles remained for examination, but they were based on 

 the year 1812 for the cod instead of 1817, which latter was the basal 

 year for the mackerel data. 



The results are given in tables ii and 12 and in figure 27. As a 

 46-year cycle had frequently been encountered in weather data, I took 

 a mean of the first, third, and fifth 23-year cycles separately from the 

 mean of the second and fourth for both mackerel and cod fisheries. 

 As there is little definite support for a 46-year cycle in these curves, 

 I also took the general mean in each case. Thus three curves for each 

 fishery are given in figure 27. 



The general mean range during the 23-year cycle for the mackerel 

 fishery is astonishingly large, from 16 to 40 millions of pounds. For 

 the cod fishery it is from 460 to 570 millions of pounds. The constitu- 

 ent cycles, as indicated by the curves of partial means, support the 

 general mean very well. Also when a difference of phase of 2 years 

 and a difference of percentage amplitude of variation are both allowed 

 for, as shown in figure 27, the two general mean curves are sur- 

 prisingly similar. As noted above, it will be observed that neither the 

 mackerel nor the cod curves show sufficient dissimilarity as between 

 the partial mean curves to prove definitely that a 46-year period is 

 superposed upon the 23-year period. Yet there are some indications 

 of it, as seen in the tendency to opposition at certain years of the cycle, 

 contrasted with the general fair agreement between the partial means. 



Table ii. — ^s-y<^<^'' Cycles in North Atlantic Mackerel Fisheries, 1817-1931. 

 Values Given in Millions of Pounds 



Cycle 



of cycles of cycles 



6 7 41 33 18 22 20 21 



7 8 37 36 6 17 22 10 



14 10 42 18 9 22 14 19 

 16 9 38 12 9 21 II 17 



15 II 33 9 12 20 10 16 

 24 28 29 21 24 24 24 24 

 20 25 36 24 24 27 25 26 

 26 34 SI 27 30 36 30 34 

 35 41 39 24 39 38 32 36 

 22 29 28 12 18 2i 21 22 

 26 32 29 36 15 23 34 28 

 32 45 42 9 18 31 27 29 

 31 25 20 9 9 20 17 19 

 43 18 35 12 12 30 15 24 

 S3 18 18 48 33 35 3i 34 



30 29 30 21 48 36 25 32 



3S 24 47 24 72 51 24 40 



26 19 53 21 60 46 20 36 



22 14 52 24 48 41 19 32 



18 32 31 12 72 40 22 33 



14 26 65 27 66 48 26 40 



10 33 38 24 69 39 28 35 



• These two columns are three times their originals, as stated in the te.xt. 



