70 



CLIMATIC CYCLES AND TREE GROWTH 



attention, 8.5 and about 14 years. They show well by directions of align- 

 ment of increased amplitudes as in Plate 15B. 



Centers of Mass of Maxima — Cyclogram analysis distinctly uses the mass 

 of maxima in producing its cycle effects. In the case of well-defined maxi- 

 mal masses, as in the sunspot cycle, the use of centers of mass is well indi- 

 cated, as in the plate referred to. This use of the centers of mass simplifies 

 the general pattern and shows the various lesser cycles, such as 14 years 

 and probably 7.0, 8.5, 10, and also suggestions near 17 and 20. The setting 

 near 23 in Plate 15B shows a symmetrical arrangement of the maxima and one 

 is less conscious of the irregularities in timing near 1800 that looked so large 

 in the 11-year setting. The mean departures from a true period after all 

 are only two to three years, say 20 to 30 per cent of the cycle length, and on 

 a 23-year setting the percentage reduces to 10 or 15. 



Possible Half-Sunspot Cycle — In the analysis of the smoothed annual 

 means we do not readily see any half cycle of about 5.7 years, such as found 

 in terrestrial phenomena and reported in radiation data (by Abbot). The 

 form of the sunspot cycle since 1750 with its bulky maxima and very low 



1902 1905 1910 1915 



Fig. 30— Monthly sunspot curve to show reduced amplitude at minimum (after 

 Wolfer). Note: This gives a striking illustration of a poor smoothing method 



minima gives apparently no indication of a secondary crest in the 11 years. 

 One finds it hard to look upon the slight hump that sometimes occurs dur- 

 ing decreasing spot activity as any substitute for a secondary maximum 

 that would nearly divide the whole cycle length in halves. A secondary 

 maximum of activity in the monthly values would be hard to find owing to 

 difficulties in evaluating the fluctuations during the long low minima, yet, 

 in spite of that, these monthly values seem so far to be the best direct solar 

 data in which we have a chance of finding a 5.7-year period, approximating 

 the half cycle in length. 



We find, however, in the tree-ring records extensive sequences of a 5.7- 

 year recurrence which we have called the Hellmann cycle; we find also 8.5, 

 10, 14, 17, 19 or 20, and 23 years, as will appear later. We find also in trees, 

 in the last 1000 years, certain cycles close to 12 years in length (see page 107). 

 Here is, therefore, the place to ask the question : Can we regard such lengths 

 as variants of the well-recognized 11-year cycle? We can not answer yet 

 with certainty, but since sunspot numbers had a limited cycle life at 14 

 years (1788-1830), we know no reason why 12-year lengths should not occur 

 in the sun. 



