66 Mr. W, Ellis. Relation between Diurnal Range of 



January, to adding together half the sum of the numbers for the 

 preceding and following July, and the sum of the numbers for the 

 intervening eleven months, August to June, and dividing the whole 

 by twelve. These new monthly numbers, each expressing an annual 

 mean, are given in Table II, both for declination and horizontal 

 force, and they are used with those of Table II of the previous paper 

 to form the two lower curves of the accompanying plate. I remarked 

 in my previous paper that the indications of vertical force were for 

 the present purpose not very manageable ; several different instru- 

 ments had been employed, and the results presented anomalies. 

 Certain beneficial alterations were, however, made in 1882, in the 

 instrument still in use, since which time it has worked better. It 

 showed a maximum diurnal range in 1883, the descent to a minimum 

 in 1889, and the subsequent rise to a maximum, although there 

 remains still some degree of irregularity of action. 



As regards sun-spot frequency, Dr. Wolf's monthly values, as 

 derived directly from observation, are given for the years 1841 to 

 1877, in Table VI of my previous paper. Those for the years fol- 

 lowing 1878 to 1896 are to be found in different numbers of his 

 ' Astronomische Mittheilungen,' the values in the later years, after 

 the death of Dr. Wolf, in 1893, having been similarly prepared by 

 Professor Wolfer, his successor at Zurich. I am not aware that 

 these have been before given in a collected form; they will be found 

 in the annexed Table III. For the purpose of smoothing the acci- 

 dental irregularities of these observed sun-spot numbers, Dr. Wolf 

 treated them in the same way as the numbers of our Table I (expressing 

 magnetic range) were dealt with to form those of Table II. Though 

 the process was here employed for a reason different to that which 

 rendered its application necessary in the case of magnetic range, the 

 similarity of treatment happily makes the resulting numbers strictly 

 comparable with the magnetic numbers. The smoothed sun-spot 

 numbers, from 1841 to 1876, June, are to be found in a table con- 

 tained in Dr. Wolf's paper, "Memoire sur la Periode commune a la 

 Frequence des Taches Solaires et a la Variation de la Declinaison 

 Magnetique."* Those from 1876, July, to 1896, added in our 

 Table III, have been in part taken, from the ' Astronomische Mit- 

 theilungen ' and in part calculated from the observed numbers con- 

 tained in the same table. These smoothed values, with those of the 

 preceding series taken from the paper above mentioned, are used to 

 form the upper curve in the diagram of collected sun-spot and 

 magnetic curves. It may be asked why the Greenwich magnetic 

 variations are not compared with the Greenwich sun-spot record. 

 But this record having been maintained only for some twenty years, 

 it was deemed better to adhere throughout to the long Wolf series 

 * ' Memoirs of the Royal Astronomical Society/ vo 43, p. 199. 



