8o 



NA TURE 



[May 26, 



present year, which appears in the Proceedings of 

 the Society. The results here employed extend from 

 1 84 1 to 1896, a period of fifty-six years. The addition of 

 the more recent observations is especially interesting as 

 contrasting in some respects with the earUer portion, the 



Fig. I. — Smoothed curves of sun-spot frequency (Wolf), compared with 

 corresponding curves showing the variation in diurnal range of the 

 magnetic elements of declination and horizontal force from observations 

 made at the Royal Observatory, Greenwich. 



whole forming one continuous chain of evidence that 

 much strengthens the argument for relation between the 

 two classes of phenomena. The records of horizontal 

 magnetic force, as well as those of declination, are 

 empJIoyed. From 1841 to 1847 the results depend on eye 



NO. 149 1, VOL. 58] 



observations made at intervals of two hours. In 1848 and 

 afterwards, they depend on hourly tabulations from the 

 photographic records. The mean diurnal range in each 

 month is taken to represent (relatively to other months) 

 the magnetic energy of the month. By the mean diurnal 

 range of declination or horizontal force is to be under- 

 stood the difference between the least and the greatest of 

 the mean hourly values in each month. 



In any graphical representation of unexplained phe- 

 nomena It IS important to give ready reference to the 

 numerical data employed. Consequently in both papers 

 complete tables of the elements used are either included, 

 or indication given where such collected results can be 

 found. The numbers for declination are in minutes of 

 arc, those for horizontal force are in parts of the force, 

 taken as unity. There being in the numbers a stronglv 

 marked annual period (the summer values being greater 

 than those for winter), numbers that shall be free of 

 annual inequality have (as explained in the paper) been 

 prepared and used to construct the middle and lower 

 curves of the diagram of collected curves. For sun-spot 

 frequency the numbers published for so many years by 

 Dr. Wolf, and since his death continued by his successor, 

 Prof. Wolfer, have been employed. It is impossible 

 to value too highly work of this kind, carried on for so 

 many years on one fixed plan ; such steady adherence 

 to a definite method having many advantages. The 

 monthly sun-spot numbers show considerable irregu- 

 larities which Wolf smoothed by a process similar to that 

 employed to free the magnetic numbers from the annual 

 inequality, the resulting numbers being used for the 

 upper curve of the diagram. 



The collected curves show striking points of interest. 

 The epochs of the extreme points of the curves are given 

 in the following table : — 



Tad/e of Epochs of Magnetic and Sun-spot Minima and 

 Alaxima. 



Mean excess (five epochs of minimum) 

 Mean excess (five epochs of maximum) 

 General mean excess 



+0-32 

 + 0-04 

 +0-18 



— O-32I Q-OO 

 -|-O'22[+0*I3 

 -0-051-1-0-06 



The intervals between the successive mean magnetic 

 epochs and the corresponding sun-spot epochs run, it will 

 be seen, closely together. And if instead of successive 

 intervals we take successive periods, as from No. i to 

 No. 3, No. 2 to 4, &c., of the table, we have— 



Length of Magnetic Period. 



12-55 



2-4 

 11-85 



3-5 



4-6 



y- 



5-7 6-8 7-9 s-io 



y. y. y. y. 



11-40 10-45 ""SO I3"05 io'90 985 

 Length of Sun-spot Period, 



y. y. y. 



12-50 I2-00 1 1 -20 



y- 

 10-50 



y- 

 13-40 



y. y. 



I -20 10 -OO 



