CYCLOGRAM ANALYSIS 



37 



In the diagram 18d we have reached the plot we have been seeking, but 

 the original ordinates of figure 18a, consisting of sunspot numbers, become 

 meaningless in the new position and are therefore swung up into a vertical 

 position perpendicular to the page, as in figure 19a, and must be expressed as 

 a third dimension, which is easily done by representing light intensities by 



40 

 I 



^nTlTnw ^rTTTT>v^/fnTrK ^fmrrh^fmTT^ ^TTTTrv^<lTTrT>>^^Tl 



18.50 



60 



70 



i 



80 

 L_ 



90 



1900 



10 

 i 



20 

 i 



30 

 I 



CL Common plot of sunspot numbers 



io 2oYears 



1840 



34 >w^fni TTTT>> 



1850 



TTT^^TTrh-^ 



I860 



TTTN^TTTTn^ 



1870 



Tn^Trt^prrffTT^ 



I88Q ^TTTm>> .^ffnTTTr ^. 



', 890 ^rTTTTr>^iTn"TT>^ 



19 10 ^ -^<nTT>^^r<nTTi 



900 ^fTrTTrr> — ^mTTT^ 



i 920 |v> ^TTfJTTh 



o 10 20 



Years 



The same divided into 

 lOyear blocks 



</> Years 



C Last figure changed to 

 rectangular coordinates 



<f)±L 

 *E 



1850 60 70 80 90 1900 10 20 30 



Time scale 

 ci Last figure rotated to bring chief 

 time scale horizontal 



Fig. 18 — a, b, c, d. Cyclogram demonstration. 



local enlargement or by some half-tone process as in common photographs. 

 Thus all maxima in the original data are retained in the plot. 



This is crudely represented in figure 19a. Here, as stated above, we 

 have the main time scale horizontal and a vertical time scale consisting of 

 residuals from 10-year multiples. We note here that if we take some date 



