A study on the secular variation of magnetic elements based on data forD,I, andHfor Oslo, 1820-1948

An attempt is made to get a general view of the secular variation of the most important magnetic elements for as long a time as possible. Those elements are D, I, H, X, F, Z, and F. For D the annual data are given for the entire interval 1770–1948 and for the rest 1820–1948. It is quite clear that at a station where, since 1890, modern photographic registering instruments are mounted, more exact data are obtained for the magnetic elements, especially where details are concerned, but also here the homogeneity might be questionable, because some of the most important observatories have been forced to change from one place to another (often considerably removed from the original site). There is also to be considered the question of the quality of the absolute observations used for the reductions. It must be remembered that there has been opportunity to compare standards only in recent years. A change of instrument and change of place, where the absolute observations were taken, may influence the homogeneity and the larger gaps may also be serious. Comparing the Oslo material with that of other stations, one gets the impression that the Norwegian series—thanks to Hansteen's imposing work—is the most complete and not worse in regard to homogeneity than other old series. The relative H-instrument has continued unaltered and uninterrupted from 1843–1930, and the reduction is based on absolute observations taken on the same pier during all these years and with Hansteen's standard magnet—most of the time by Hansteen personally. In D the conditions are not so favorable, but all are fairly good. The I-series is probably also fairly homogeneous between 1870 and 1908, but there is a deplorable gap between 1908 and 1948. No registrations are at hand before 1870 for I and before 1843 for D and H. Nevertheless, annual data for D from 1770 and for H and I from 1820 have been found in several series of absolute observations. The technique used for interpolation and extrapolation is fully explained. The most serious gap is the one for I during 1908 to 1948, but also in this case the results are not far from true. As to Hansteen's absolute observations, one is impressed by the quality he actually obtained with the primitive instruments he used.