APPENDIX 165 



in this figure correspond to a period of 150 days, 160 days, etc. There are 

 two places in the figure where the corrugations are especially evident; one 

 had about 157 and the other had about 187 days. In the instrument the 

 latter is immediately recognized as the true period, because the lines are 

 alternately yellow and white. 



Plate 24 shows a differential pattern of SS Cygni, so taken that vertical 

 lines of dots show a period of about 110 days. The predominant periods 

 may easily be worked out from this diagram. For the first four years, begin- 

 ning about 1896 the double-crested 117-day period was the most prominent. 

 Then for six years a double-crested asymmetrical period of 107 days was the 

 most prominent. Then there was nearly a year of readjustment followed by 

 two years of 110-day double-crested periods. This was followed by three 

 years of 125-day double-crested periods, returning finally to four years of 

 117-day double-crested asymmetrical periods. This alignment of dots is 

 best seen by looking at the figure from a low angle rather than from the 

 perpendicular position of ordinary reading. 1 



1 In considering this pattern eighteen years after it was first made and comparing 

 it with cyclograms of solar records and of climatic data, we note that it shows the type 

 of persistence, of bisection and of somewhat irregular changes that appear, for example, 

 in the magnetic character figure C. In this analogy we have a suggestion that this 

 type of variation might have its origin in the rotation periods of a large body like 

 our sun. 



