THE SUN-SPOT PERIOD. 



145 



the 3'eai" in question. For brevity denote the resulting temperature 

 deviations b}^ <^^ and the sun-spot numbers by SS. We thus obtain 

 Table V. 



Table V. — 3raxima and minima of san spots. 



Comparing the values of (^S with the numerical values of (J^, it 

 appears that they vary in general in the same sense for years of max- 

 ima of sun spots, and in opposite senses for years of minima. But if 

 we regard algebraical signs, a maximum maximorum of (JS corre- 

 sponds to a minimum minimorum of S6^ and vice versa. 



5. Finally, the principal points of the preceding discussion may be 

 implicitly summed up as follows: 



The function of temperature departures which we have just con- 

 sidered ma}' be referred to a new origin of ordinates such that the 

 departure +0'-.25 becomes the new zero; and we may reckon the new 

 ordinates in the direction which was formerly that of increasing nega- 

 tive departures. Denoting by 36 the new ordinates as thus consid- 

 ered, it Avill be seen that Sd represents in some sort (other things being 

 equal) the difference for each year between the temperature which 

 would have been experienced if there had been no sun spots, and that 

 which was experienced in realit}'; for the origin of ordinates at +0^.25 

 corresponds with the conditions of the year 1900 when there were 

 scarcely any sun spots. If now we take the mean of the values of Sd 

 for the lirst half period 1870 to 1881, and multiph' this mean by the 

 number of years in this half period, and deal similarly by the periods 

 1881 to 1889, and 1889 to 1900, also treating the sun spot frequency 

 data (^S after a similar fashion, the results obtained are expressed b}^ 

 the following table. ^' 



Table VI. — Sumviatlon of the temperature curves and of the inverse sun-spot frequency 



curves. 



«The unit of temperatures for tlii.s table is the luindredtli of a degree eeutijjrade. 



