1 66 



NATURE 



[Dec. i 8, 1870 



SUN-SPOTS AND THE RAINFALL OF PARIS 



IN a paper on this subject by Mr. C. Meldrum, F.R.S., read 

 at the Meteorological Society of Mauritius, after some preli- 

 minary remarks, the author said : — 



The rainfall observations made at the Observatory of Paris 

 formed perhaps the longest series on record. They were com- 

 menced in 1689, and, with the exception of twenty-six years, 

 viz., 1697-98, 1755-72, and 1798-1803, they had been continued 

 to the present day. 



From 1689 to 1870 there had been, according to Dr. Rudolph 

 Wolf and Prof. Fritz, seventeen years of maximum and seventeen 

 years of minimum sunspot. Now it would be seen from the 

 following table showing the years of maximum and minimum, 

 and (as far as possible) the rainfall in each of them at the Paris 

 Observatory, not only that more rain had fallen in the former 

 than in the latter, but that throughout that long period there had, 

 as far as could be ascertained, been only two exceptions to the 

 rule that the maximum were wetter than the minimum years. 



From the above table it followed : 



1. That the mean rainfall of the thirteen minimum sunspot 

 years in the second column was 454-6 mm., and of the fifteen 

 maximum years in the fouith column 524-3 mm., giving a mean 

 annual excess of 69-7 mm. in favour of the latter. 



2. That the results in the fifth column, of direct comparisons 

 of the rainfalls in thirteen minimum with the rainfalls in thirteen 

 maximum years, gave a mean excess of 67-7 mm. in favour of the 

 latter. 



3. That comparing the rainfall in each minimum year, from 

 1 689, with that of the following maximum year, there were only 

 two minimum years (1712 and 1S67) in which the rainfall had 

 not been less than in the maximum year. 



4. That comparing the rainfall in each maximum year, from 

 1693, with the rainfall in the following minimum year, there was, 

 as far as was known, only one maximum year (1705) in which 

 the rainfall was not greater than in the minimum year. 



5. That, as a rule, therefore, the rainfall of a maximum fun- 

 spot year was greater than that of either the preceding or follow- 

 ing minimum year ; a circumstance which seemed to indicate a 

 tendency, at least, to a periodic variation in the rainfall of raris. 



The most important feature was, not lhat on the whole the 

 rainfall of the maximum had exceeded that of the minimum 

 years, but that the excess had occurred in eleven out of thirteen 

 cases. This frequent repetition of the same phenomenon pointed 

 to a periodicity. A mere excess of rainfall in the maximum 

 years, as a whole, could not have done so ; for such an excess 

 might have happened if only a much smaller number of the 

 maximum years had been wetter than the preceding or following 

 minimum years. But that was not the case. From 1723 down 

 to 1867 there was not, as far as the observations went, an 

 instance in which the rainfall of a maximum year did not exceed 

 that of the previous and next minimum year. 



To the possible objection that the years of maximum and 

 minimum sun-spots of the seventeenth and eighteenth centuries 

 had probably not been so well determine.! as those of the nine- 

 teenth century, it could be replied that the results for the nine- 



teenth century, also, showed that the rainfall had been greater 

 in the maximum than in the minimum years. From 1804 to 

 1867 there had been six maximum and six minimum years. 

 Now in every case the rainfall of the former exceeded that of 

 the latter, and the mean excess was 86" 1 mm. 



Since 1867 there had been only one maximum year, viz., 1870, 

 and the next minimum year was not yet known. It was true 

 that the rainfall of Paris in 1870 had been comparatively small, 

 and it was not improbable that it would be less than that of the 

 next minimum year ; but if such should be the case it would be 

 the only exception to the general rule since the commencement 

 of the century, if not since 1705. 



The total rainfall at the Paris Observatory in the seven 

 maximum years, since 1800, was 4,004-2 mm., and in the six 

 minimum years 3,069-5 mm. A rainfall of 934-7 mm., therefore, 

 would be required in the seventh minimum year to restore the 

 balance ; and there was very little chance of this, the greatest 

 recorded rainfall, since the observations had been commenced, 

 having been 703" 1 mm - m the maximum year 1804, and the 

 least 2IO"2 mm. in the minimum year 1733. 



The average duration of the sun-spot cycles was, according to 

 Wolf, I fi years. The last five complete cycles, starting from a 

 minimum year, were from 1810 tc 1867. Taking in each of 

 these cycles the three years ;of most and the three years of 

 fewest spots, and comparing the rainfall of Paris in the former 

 triennial periods with that in the latter, it was found that the 

 rainfall in each minimum period was less than that in the 

 following maximum period. The figures were as follows : — ] -jj 



It would be seen, however, that the rainfall in the minimum 

 period 1865-67 exceeded that in the maximum period 1870-72 of 

 the cycle which had commenced in 1867 (and the end of which 

 was not yet fully known), the rainfall of 1866 having been 

 abnormally great. But as the rainfall in the three years 1871-73 

 (iSi7'9 mm.) when the sun-spots were still numerous, had 

 exceeded that in the three years 186567, the usual excess in the 

 maximum periods may have only been somewhat retarded. 



By forming, in the manner described on previous occasions, a 

 mean cycle out of the five cycles from 1810 to 1867, and com- 

 paring the rainfall of Paris with the sun-spots for each year, the 

 follow ins; mean results were obtained : — 



The above table showed that both the spots and the rain were 

 above ( + ) or below ( - ) their respective averages in the same 

 years of the mean cycle, except the last, and that they both 

 attained their maximum in the same year, namely, the fifth. 

 The discrepancy, with respect to rtlpi, in the eleventh year, 

 was owing to the years 1854 and 1866 having been abnormally 

 wet. 



From the maximum year 1816 to the maximum year 1870 

 there had also been five complete cycles, which, omitting 

 fractions, gave the following results : — 



