7'2 Mr. W. Ellis. Relation between Diurnal Range of 



consequently on many that are retained considerable disturbance 

 exists. Now abnormal disturbance is more frequent, as well as 

 greater in magnitude, towards the epochs of maximum sun- 

 spot frequency, and tends almost to disappear at the epochs of 

 minimum frequency. For instance, the sun-spot minima of 1856 

 and 1879 were both remarkable for little irregular disturbance, 

 especially the latter, whilst at the sun-spot maximum of 1870, dis- 

 turbance was unusually frequent and considerable in amount, the 

 difference in this respect being very striking, as a mere inspec- 

 tion of the photographic records for the epochs mentioned would 

 abundantly show. In preparing the former paper the question, 

 therefore, did arise whether and to what extent the increased diurnal 

 magnetic range, about the periods of sun-spot maximum, might be 

 due to the greater prevalence at such periods of abnormal disturb- 

 ance, but I then satisfied myself, by examination of the records, that 

 this circumstance exercised no important influence on the results. 

 Now, however, I can substantiate that conclusion by numerical data. 



It is known to those acquainted with magnetic work that in order 

 the better and more readily to compare together the diurnal magnetic 

 inequalities for different places it has been the practice, since the 

 year 1889, to tabulate the records at British observatories for five 

 selected quiet days in each month, the selection of days being made 

 by the Astronomer Royal. These quiet day results, unaffected by 

 magnetic disturbance, show only the solar diurnal variation, and are 

 now available for Greenwich, from 1889 to 1896. Mean hourly 

 values being formed in each month, from the indications on the five 

 quiet days, the diurnal range of declination and horizontal force was 

 found in the same way as before described for the full monthly 

 values, the resulting numbers being given in Table IV. These were 

 further treated for removal of the annual inequality, in the same way 

 that the numbers of Table I were treated to form those of Table II. 

 The values so found are contained in Table Y, and are those repre- 

 sented graphically in fig. 3, adding thereto the graphical representa- 

 tion of the sun-spot and magnetic phenomena, for the corresponding 

 years from the long series 1841 1896, in which all days (excepting 

 those of extreme disturbance) were included. 



The thick line in declination and horizontal force (fig. 3) shows the 

 diurnal magnetic range as found from employing all days, including 

 many of magnetic disturbance. The thin line, formed from five 

 selected qniet days in each month indicates the true solar diurnal 

 range. Two things thus appear. Firstly, that the solar diurnal 

 range (thin line), in which the influence of irregular magnetic dis- 

 turbance has no part, is itself really affected with a periodical varia- 

 tion similar to that of sun spots. And secondly, that the effect of 

 including disturbed days (thick line) alters the solar diurnal range 



