J. P. Hall— Short Cycle in Weather. 239 



August, he declared that he felt able to show that a real rela- 

 tion does exist between magnetic and meteorological phe- 

 nomena. 



This assurance and the facts adduced in the present article 

 suggest the following questions : 



(1) What is the sun's exact synodical rotation-period? This 

 needs to be determined accurately before we can make it the 

 basis of comparison with terrestrial phenomena. The testi- 

 mony of spots, magnetic storms, auroras, and spectroscopic 

 observation of the sun's motion at the limbs is conflicting and 

 diverse, yielding results all the way from 26 to 29 days. The 

 mean time required for spots to reappear on the solar disc has 

 been estimated at the Greenwich Observatory to be 27^ 

 days; but as yet we are not justified in regarding even this as 

 final or representative of the whole orb. 



(2) If the periodicity in terrestrial weather be due to the 

 return, by the sun's rotation, of some important solar feature, is 

 that feature a spot, facula or prominence, liable to develop 

 in any longitude, and seldom existing for much more than 

 three or four rotation-periods, or is it a permanent center or 

 site? In the former case, the recurrence of any particular 

 weather incident could only be counted upon a few times ; in 

 the latter it might be predicted not only months, but even 

 years in advance. In the traces here reproduced, for instance, 

 a warm wave, like C in Series I, will appear, without any 

 immediate precedent, and subside after a few appearances. 

 In some such cases purely terrestrial and local interferences 

 can be discerned ; in others, the reason for subsidence is not 

 apparent. This would seem to favor the notion that a short- 

 lived solar disturbance was the agent. On the other hand, 

 the traces in Series I represent a minimum of the 11-year 

 sunspot cycle, while Series II represents a period, which, 

 according to Dr. Rudolph Wolf (whose "sunspot numbers" 

 are widely accepted as a standard), exceeded in evidence of 

 solar activity the last maximum stage.* Yet the periodicity 

 of the temperature-changes is about as distinct and the ampli- 

 tude nearly or quite as great, in one series as in the other. 



(3) If the exciting influence is associated with one or more 

 permanent sites, is the occasional subsidence and revival ob- 

 served in its effects due to variations in the energy itself, to 

 unequally advantageous positions resulting from the inclina- 

 tion of the sun's axis, or to some unequal terrestrial masking 

 or interference? Perhaps an answer to this can be found 

 more easily when the following question is disposed of : 



*The mean of Wolf's monthly numbers from Sept. 1, 1889, to March 31. 1890, 

 is 3-5; for the year 1883, 62-8; for 1884, 63'3 : for the first nine months of 1892, 

 74-0. 



