On the occasion of a brilliant display of aurora some apparently rotatory movements in the position of the corona suggested to the author the possible existence of similar changes of direction in the total magnetic force during disturbance. For investigating these changes the most convenient available data were the plates published with the “Greenwich Magnetical and Meteorological Observations” (hereafter referred to as the “Report”), showing on a reduced scale the diurnal registers of the three components of magnetic force for several disturbed days in each year. The eight years 1900-7 were dealt with, and a preliminary inspection of all the registers figured for these years was made to select an example for special consideration. The published registers are reduced in scale from the originals in the ratio 11:20, and to lessen the chance of error on this account, it was decided to select a disturbance showing comparatively slow, but steady and decided, changes. The registers selected in the first instance were those for October 12, 1903, and it was decided to determine the variation in direction of the total force from 18 hours onwards, at intervals of about 5 minutes, leaving out of account the effect of the small temperature variations recorded, for reasons presently to be stated. This variation in direction was found in terms of transverse deflecting force, the unit employed being 1 x 10
-5
C. G. S. (1
γ
). In the following remarks the letters D, H, and Z denote the declination, the horizontal force, and the vertical force, respectively; F and I, respectively, denote the total force and the dip; while the prefix ∆ signifies inequality increments. Ordinates of the components D, H, and Z were measured on the plate at time intervals of 1/40 inch, corresponding to a measure once every 5 minutes, approximately. The zero of a scale divided to 1/40 inch was applied to the immediately preceding hour line, and points on the respective traces were thus set off at equal increments of time, and, from these, the force ordinates were measured to the adopted base lines. A division of the hour intervals into 12 equal parts was not so important as having the ordinates measured at exact time equivalents in all three registers. Regarding the accuracy with which the three registers were placed in time juxtaposition for the purpose of photo-lithographic reduction there is the assurance given in the “Report” (p. 37 for 1903) that great attention has been paid to this detail.
Small temperature variations are a key regulator of reproductive growth and assimilate storage in oil palm (Elaeis guineensis)
