A REPRESENTATION OF THE SUNSPOT CYCLE 293 



and ^ is a constant depending on the type of telescope and other 

 factors affecting the observation. The figure 10 is an arbitrary one 

 arrived at by Wolf from investigation of a number of individual cases 

 and which seemed to give him the proper relationship. 



A careful study was then made by Wolf of all existing records of 

 prior data. Hofrat Schwabe of Dessau supplied data for the period 

 1826 to 1855 to which a correction was applied determined from a 

 study of the overlapping data and from the percentage of spotless days. 

 Johann Casper Staudacher of Niirnberg had made a total of 1131 

 observations (from one to ten every month) by means of a helioscope 

 during the period 1749 to 1799. He often gave detailed descriptions 

 and included many sketches. Imagine a man making observations for 

 fifty years without any attempt at analysis and then have the data 

 resurrected fifty years after his death to form an important con- 

 tribution to the record. Flaugergues (1794-1830), Tevel (1816- 

 1836), Adams (1819-1823) and Arago (1822-1830) supplied most of 

 the data for the intervening period between the observations of 

 Staudacher and Schwabe. Wolf lists about a hundred references 

 (225 up to 1866) to sunspots prior to 1850. In most cases, the obser- 

 vations were incidental to other solar observations such as culmina- 

 tions, solar diameter, eclipses, transits, or on the nature of sunspots 

 rather than their number. Each record was carefully studied, and 

 from them all Wolf obtained a representation of sunspot numbers 

 for as far back as 1749 and established the times of maxima and minima 

 with an accuracy of ± 2 years or better back as far as 1610 A.D. 

 Although the data prior to 1849 include a certain element of un- 

 reliability and all the data represent relative numbers which have been 

 obtained from the observations by applying a weighting factor, it is, 

 however, not only the best record but the only one for such a long 

 period of time. In the aggregate it is probably a good indication of 

 the variation of solar activity. 



The method employed in the present analysis is briefly as follows : 



(a) The yearly averages of sunspot numbers from 1749 to the end 

 of 1937 were first plotted in the conventional way as shown in Fig. 1 ; 

 it was noted that in certain sections of the curve, notably after 1840, 

 the maximum amplitudes of alternate eleven-year periods were 

 higher than the intervening ones. 



ih) The data were redrawn with alternate eleven-year periods above 

 and below the time axis; this not only smoothed out the envelope 

 of the maxima but also simplified the analysis by eliminating a com- 

 puted mean value base line which has been employed in previous 

 analyses; the maximum-amplitude component becomes approximately 



