June 20, 1901] 



NA TURE 



197 



The sun may thus be considered a variable star (i) the light spotted area from minimum to niinimum is least (curve D), the 

 of which (reckoning from minimum to minimum) is variable, long period rainfall curve is at its maximum or we have a wet 

 with a mean value^of about ifl years; (2) the epoch of I cycle. When, on the other hand, the maximum (curve b) follows 

 maximum does not occur a constant number of years after the j soon after the preceding minimum (1867), and the spotted area 

 preceding minimum, but varies regularly, the cycle of variations for this cycle is at a maximum (curve D), the rainfall curve is at 

 covering about thirty-five years. ' a minimum or a dry cycle is in progress. 



It is interesting, therefore, to know that the sun is not the only I Prof. Ed. Richter, in a detailed investigation of the movement 



star which exhibits variations similar in kind to those mentioned 

 above, for the light curve of r/ Aquila: not only has a more rapid 

 rise to maximum and slosv fall to minimum, but the periods 

 from minimum to minimum vary in length, and the interval 

 minimum to maximum has a regular variation. 



Since then, in addition to the well-known eleven-year period 

 of sunspot frequency, there is another cycle which extends over 

 about thirty-five years, and which is in- 

 dicated clearly, as has been shown, both 

 by the changes in the times of the occur- 

 rence of the epochs of maxima and in the 

 variations in area included in consecutive 

 eleven-year periods of both sunspot and 

 magnetic curves, it is only natural to sup- 

 pose that this long-period variation is the 

 effect of a cycle of disturbances in the 

 sun's atmosphere itself. 



Such a cycle, if of sufficient intensity, 

 should cause a variation from the normal 

 circulation of the earth's atmosphere, and 

 should be indicated in all meteorological 

 and like phenomena. 



We are indebted to Prof. Ed. Briickner 

 for the great work on the changes in 

 climates, and in this investigation he 

 sought variations in the observations of 

 the height of the waters in inland seas, 

 lakes and rivers ; in the observations of 

 rainfall, pressure and temperature ; in 

 the movements of glaciers ; in the fre- 

 quency of cold winters ; growth of vines, 

 &c. 



The result of the whole of the investi- 

 gation led him to the conclusion that there 

 is a periodkal variation in the climates 

 over the whole earth, the mean length of 

 this period being 34 'S + 07 years. 



Prof. Bruckner was so convinced of the 

 undoubted climate variations which he 

 deduced, and so certain that such variations 

 could only be caused by an external influ- 

 ence, that he investigated Wolfs sunspot 

 numbers to see whether such a cycle was 

 indicated. Not finding any he was led to 

 make the bold suggestion that such a 

 variation as he sought must really exist in 

 the sun, but might possibly be independent 

 of sunspots. He finally concluded that 

 the climate variations are the first symptom 

 of a long period variation in the sun, which 

 probably will be discovered later. 



In the light of the secular period of 

 solar activity dealt with in this ariicle. 

 Prof. Bruckner's conclusions are of great 

 interest, because not only does the length 

 of the period, but the critical epochs of 

 his cycle completely harmonise with those 



found in the present discussion of the sunspot and magnetic 

 curves. 



To illustrate more fully this connection and to take only 

 one case, namely, rainfall, three rainfall curves which have 

 been copied from his book are reproduced in Fig. 2 (curves 

 B, F, G). 



E and F represent the secular variations for what Prof. 

 Briickner calls " Regulare Gebiete I and II," while curve E is 

 the mean for the whole set of observations he has employed, 

 and represents the secular variation of rainfall over the whole 

 earth so far as can be determined. 



The comparison of these curves with those representing the 

 sunspot and magnetic results given above them shows that when 

 the epoch of maximum spotted area (curve b) follows late after 

 the preceding epoch of minimum (1S43, 1878), or when the 



of glaciers, has also found a cycle of thirty-five years, and he 

 pointed out that the variations agreed generally with Bri'ickner's 

 climate variations, the glacier movement being accelerated during 

 the wet and cool periods. 



Again, Mr. Charles Egeson not only finds a secular period of 

 about thirty-three to thirty-four years in the occurrence of rain- 

 fall, thunderstorms and westerly winds in the month of April 



8^00- 



zpoo- 

 IfiOO- 



ipoo- 



SUNSPOT 



CURVE. 

 (YEARLY MEANS.) 



SUNSPOT 



CURVE. 



MAGNETIC 

 CURVE. 



TOTAL SUNSPOT ^ 

 AREAS, 



[PER PERIOD) 



ERUCKNER''5 

 CLIMATE 



VARIATIONS. 



SAINFALL. .yQ 



/. REGULARE QEBIETE.I. 



. V^HOLE EARTM. 



for Sydney, but the epochs of maxima of the two latter 

 harmonise with the epochs of the thirty-five yearly period 

 deduced for sunspots. 



There seems little doubt that, during the interval of time 

 covered by the present sunspot discussion, the meteorological 

 phenomena, number of auror.x, and magnetic storms show 

 secular variations of a period of about thirty-five years, the 

 epochs of which harmonise with those of the secular variations 

 of sunspots. As we are now beginning to approach another 

 maximum of sunspots which should correspond both in intensity 

 and in time of occurrence after the epoch of the present 

 minimum with that of iSyo'S, it will be interesting to observe 

 whether all the solar, meteorological and magnetical phenomena 

 of that period will be repeated. 



William J. S. Lockyer. 



NO. 1 65 I, VOL. 64] 



