ANALYSIS OF SOLAR RECORDS 67 



influence admission of heat to the lower atmosphere and the earth's surface, 

 but we recognize that the heat radiation from the sun is the obvious factor 

 to be studied in connection with rainfall and tree growth. Thanks to Dr. 

 Abbot and the Smithsonian Astrophysical Observatory, and others, almost 

 daily records of radiation have been obtained since 1918 and we have some 

 values going back several years earlier. These, with the results of Dr. 

 Pettit and Dr. Nicholson of Mount Wilson Solar Observatory, will be dis- 

 cussed later, together with a cyclogram analysis of the international mag- 

 netic character figure C. Various long-continued solar records have emerged 

 from these investigations and offer a favorable field for cycle analysis by our 

 method. 



ANALYSIS OF SUNSPOT NUMBERS 



Very large spots were occasionally seen by the naked eye before the in- 

 vention of the telescope in 1610, and after that these solar phenomena were 

 observed more or less continuously. After 1750 and especially since 1830, 

 they were watched with care. 1 A study of the periodicities in such records 

 by aid of cyclogram analysis is a most illuminating example of the method 

 and at the same time it is the best introduction to our problems in climatic 

 and tree-ring cycles. There are two parts to such analysis: one treats the 

 well-known annual values commonly called sunspot numbers, and the other 

 handles the monthly values, which are much more irregular. The first 

 series gives a persistent cycle averaging something over 11 years since 1750. 

 The other gives a flow of several cycles, interlacing, mixed, starting and stop- 

 ping, such as we find in climatic and ring records. We find traces of this 

 second type of cycle in the first series of common annual sunspot numbers. 

 We consider below each of these groups of data in turn and add to them an 

 examination of Abbot's and Pettit's radiation measures and the records 

 since 1906 of the international magnetic character figure C, followed by a 

 review of naked-eye sunspots and northern lights which give traces of solar 

 history centuries before the invention of the telescope. 



Analysis of Smoothed Annual Sunspot Numbers — Schwabe in 1843 found 

 that the number of spots on the sun had an ebb and flow, in an interval of 

 10 or 11 years. They had been well observed since 1750. Starting with that 

 date as a maximum, other maxima had fallen approximately at 1761, 1770, 

 1778, 1788, 1804, 1816, 1830, and 1837. These intervals are roughly 11, 9, 

 8, 10, 16, 12, 14, and 7. Here was a scattering of values that started the 

 idea of a cycle as a persistent series of random values that never depart 

 very far from a mean. But later this instability of interval quieted down; 

 since 1837 the maxima have come steadily on a period averaging 11.4 years, 

 1837, 1848, 1860, 1871, 1884, 1895, 1906, 1917, 1928: many of these dates 



1 The reader is referred to any good recent text-book on astronomy for tliscussion 

 of modern opinion about sunspots and their cycle. Our sunspot data come from the 

 Mitteilungen of Wolf, Wolfer and Brunner. The pre-telescope data come from Fritz 

 (Reed's translation in Monthly Weather Review) and Maunder in Splendour of the 

 Heavens. 



