144 



THE SUN-SPOT PEEIOD. 



periods from 1870 to 1900 were not exceptions to this rule, and it will 

 be noted that the 2^ corresponding- temperature periods also conform 

 to it, as appears in Table III. 



Table III. — Number of years elaj)smg between times of successive maxima and ininlma. 



For both sun spots and temperatures we find: 

 Interval 2-3 less than 3-4, and 4-5 less than 5-6. 



Moreover, for both temperatures and sun spots the intervals 2-3 and 



4-5 are smaller than an}^ of the intervals 1-2, 3-4 and 5-6, despite 



the inequality of the total periods embraced in the two and one-half 

 cycles considered. These include the half period of eight years (1870 

 to 1878), a full period of eleven years (1878 to 1889), and a second full 

 period of twelve years (1889 to 1901). 



3. Again, designating as "rich in sun spots'' tlie years for which 

 the Wolf relative numbers exceed 60, and as "■ poor in sun spots "' those 

 years in which they fall below 15, we may form the following table, in 

 which the temperature variations are divided between the classes so 

 defined: 



T.VBLE IV. 



This comparison also exhibits a satisfactory accord between the two 

 kinds of phenomena. 



4. Let us now consider the years of maxima and minima of sun 

 spots, and for each of these years calculate a ''smoothed'"' sun spot 

 number by taking the mean between the number for the given year 

 and the half sum of the numbers for the years next preceding and 

 next following, respectively. We proceed similarly with the temper- 

 ature deviations, thus treating both kinds of data in a way to eliminate 

 secondary influences, while leaving a preponderating importance with 



