294 Prof. Balfour Stewart and Air. W. L. Carpenter. 



Sun-spot Observations and their Treatment. 



6. We have begun by taking daily values of sun-spot areas com- 

 mencing with Schwabe's observations, which have been reduced by 

 Messrs. De La Rue, Stewart, and Loewy. These reduced records 

 giving total areas for individual days will be found in an Appendix to 

 the Report of the Solar Physics Committee. They extend from 

 1832, the date when Schwabe appears to have matured his system of 

 observation, to 1853 inclusive. From 1854 to 1860 inclusive, we have 

 made use of Carrington's accurate observations ; for 1861 we have had 

 to fall back once more upon Schwabe ; while from 1862 to 1867 in- 

 clusive we have the Kew series of observations. 



These various series form, when put together, one whole series of 

 36 years. In many cases, owing to bad weather, there are gaps or 

 blanks between observations which wo have filled up by the simplest 

 possible form of interpolation. These 36 years' observations, when 

 the blanks have been filled up, give us therefore one long series of 

 more or less trustworthy daily records of the total area of spots on 

 the sun's visible disk, the unit of measurement being the one- 

 millionth of the sun's visible hemisphere. 



7. Let us now in the first place group together these sun-spot 

 observations in rows of 24 days each, dividing the whole body of 

 observations so tabulated into 36 yearly series. A year will contain 

 generally 15, but occasionally 16 lines of 24 places each, that is to 

 say, it will consist of 15, or sometimes 16, horizontal rows, the number 

 of places in each horizontal row being always 24. The sums of each 

 of the 24 vertical columns for each year are now taken. We have 

 thus for each year a series of 24 sums. But it is obvious that these 

 sums will be greater in years of maximum than in years of minimum 

 sun-spots, so that to use them as they stand would virtually mean to 

 give great weight to years of maximum as compared with years of 

 minimum sun-spots in our search for Inequalities. We have en- 

 deavoured to overcome this difficulty in the following manner. If 

 sun-spot observations be put into a graphical form, it will at once be 

 seen that whenever the total sun-spot area is great, the oscillations 

 of this, or ranges of sun-spot Inequalities, are great also, and that 

 these ranges may, as a first approximation at least, be deemed 

 proportional to the general sun-spot activity. In other words, the 

 oscillations of areas are nearly proportional to the areas. We have, 

 therefore, taken the above-mentioned 24 numbers for every year, and 

 proportionally altered them so that the sum of the 24 shall for each 

 year be equal to 24,000. Each sum for each year should, therefore, 

 be equal to 1000, provided it be not subject to the influence of some 

 Inequality. 





