THE SOLAR HYPOTHESIS. 237 



there is a very direct and close connection between the sun-spot period and the various 

 magnetic plienomena of the earth. It is possible that these magnetic phenomena are 

 intimately connected with our cyclonic storms or other climatic factors, but little is 

 known about this and we can not here discuss it. As to the general relation of sun-spots 

 to the cormnon climatic elements of temperature, pressure, winds, precipitation, and the 

 like, the matter may well be summed up in the words of Hann, the great modern authority 

 on climate:* 



" The results of very numerous and complex investigatious of the connection of the sun-spot 

 period with variations of the meteorological elements have not wholly corresponded to expecta- 

 tions. The influence of the sun-spots on the meteorological elements has been proved as com- 

 paratively unimportant. Only in the most favorable cases is one in the position to consider that 

 the traces of a parallel course in the progress of certain meteorological elements and in that of the 

 sun-spot frequency is proven. There can be no thought of the prediction of the course of the 

 weather on the ground of the sun-spot cycle." 



From this broad general statement Hann goes on to show that the amount of agreement 

 between sun-spots and climatic phenomena varies greatly according to the part of the 

 earth and the precise climatic elements which are investigated. In general, temperature 

 is the element which shows the closest agreement with solar changes, and this applies 

 much more to equatorial than to other regions. To quote Hann once more (p. 356) : 



" For the best-grounded demonstration of a sun-spot period in the mean annual temperature of 

 the various regions of the earth our thanks are due to Koppen.f In the tropics the parallelism 

 of changes in the mean annual temperature and of the frequency of spots on the sun is compara- 

 tively well proved, in middle and higher latitudes less well. The mean amplitude of the changes 

 in the annual temperature from a sun-spot minimum to a maximum amounts within the tropics to 

 0.75° C, and beyond the tropics to 0.54° C. The course of the phenomena within the tropics 

 appears from the following numbers, which show the departure of the yearly mean temperature 

 from the mean for a long period. 



Table 13. — Sun-spot Periods in the Yearly Mean of Temperature iHlhiri the Trojncs. 



Sun-spot maximum — 0.32° 



1 year after maximum — 0.27 



2 years after maximum — 0.14 



3 years after maximum -f- 0.08 



4 years after maximum + 0.30 



.5 years after maximum + 0.41 " 



Sun-spot minimum -|- 0.33 



1 year after minimum -|- 0.15 



2 years after minimum — 0.04 



3 years after minimum — 0.21 



4 years after minimum — 0.28 



The maximum of temperature falls about 0.9 year before the sun-spot minimum, 

 while the minimum of temperature practically coincides with the maximum of the spots. 



As to the other climatic elements, the case is by no means so clear, and the results 

 sometimes appear to be contradictory. The reader who would carry the matter further 

 is referred to the last chapter of volume 1 of Hann's Klimatologie, and to the large number 

 of references there cited. In general it appears that the strongest evidence of a sun-spot 

 cycle in climate is found when a single element, such as summer rains or tropical cyclones, 

 to take the two best examples, is considered alone, and when it is investigated by the use 

 of means for a large number of stations and for long periods. WTien single stations or single 

 sun-spot cycles are considered there is likely to be no visible relation whatever. In rainfall, 

 as in temperature, there is decidedly more evidence of a sun-spot cycle within the tropics 

 than in other parts of the world. One of the most noticeable cases and one of the few which 

 is distinct and unmistakable is found in the number of tropical cyclones or hurricanes both 

 in the Indian Ocean as investigated by Meldrum and in the Atlantic according to Pocy.J 



* J. Hann, Handbuch der Klimatologie, Stuttgart, 1908, vol. 1, pp. 355-356. 



t W. Koppen, tjber mehrjahrige Perioden der Witterung in besondere iiber die ll-jahrige Periods der Temperature. 



i See Hann, Klimatologie, vol. 1, p. 360. 



