678 MR J. A. BROUN ON THE DIURNAL VARIATIONS OF THE 



perceive much greater differences in the daily movements. In January 1859 the 

 minimum occurs in the first fortnight betwixt 6 h 30 m a.m. and noon, returning 

 again to 6 h 3 m a.m. to reappear near noon at the end of the second fortnight, 

 A similar change of hour of minimum is shown in the latter half of December 

 1858, the curves for which I have also projected (Plate XLII.) The maximum 

 also happens betwixt 9 h 30 m a.m. and 4 h 30 m p.m., the curve becoming nearly 

 inverted about the 14th and 28th of January. This inversal also, it will be seen, 

 does not take place suddenly, but gradually, and without the irregularity shown 

 in disturbances in high north latitudes, as at Makerstoun. Similar, though less 

 marked, .differences of movement appear in the curves for February 1859. 



Lunar Action : sometimes as great as that of the Sun. 



This curious variation of the form of the diurnal curve is due chiefly to the 

 action of the moon. Hitherto the lunar action has been supposed so small, com- 

 pared with the solar action, that it has been concluded we might neglect the former 

 in considering the laws of the latter, and that the lunar action could be made visible 

 only in combinations of masses of observations, as a residual quantity nearly of 

 the second order. It will be found from what follows, that the lunar action is 

 sometimes as great as, if not greater than, the solar action. 



I shall consider more minutely, on another occasion, the subject of the varia- 

 tions of magnetic declination due to the action of the moon, and shall then state 

 the methods employed by me to obtain the results. It will suffice at present to 

 say, that, assuming the daily curves to be due to solar and lunar actions super- 

 posed, if we subtract from each daily curve the mean solar curve of the corres- 

 ponding month, the remainder may be supposed due to the lunar action, and to 

 other causes, irregular or regular (such as a variation of the solar action). As the 

 monthly mean curves are undergoing a gradual change, those corresponding to 

 the middle of each week were obtained, and these mean curves were subducted 

 from the daily curves in the respective weeks. The results for the lunar month, 

 December 16, 1858, till January 15, 1859 — a month, showing a marked lunar 

 action, have been projected, Plate XLIII. 



I have given — 1st, The daily curves (see Plate XLII.); 2d, The differences 

 obtained as above, under their respective solar hours ; and, 3d, The differences 

 under their respective lunar hours (see Plate XLIII.) — that is, the moon's nearest 

 hour angle when the observation happened, the latter representing the lunar 

 daily curves, with the addition of irregular or other effects.* 



These projections will explain, to a great extent, the curious variations of the 

 epochs in the solar diurnal curves which the curves of differences resemble, ex- 

 cepting that they show, more distinctly, the movement of maximum and minimum 



* As the moon takes more nearly 25 solar hours to return to the same meridian, it is assumed 

 that there are 25 lunar hours in 360 degrees. 



