Horizontal Component of Magnetic Force. 



363 



Considei'ing, first, the mean for the sixty days — being the mean for the year — we find that the minimum 

 of the horizontal component occurs, as before, at lO** 10"" a.m. ; the maximum occurs at G*" 40™ p.m., being 

 fully an hour after the epoch for the whole observations, and the secondary minimum is almost entirely 

 obliterated. The same epochs are obtained from the mean for the 120 days ; in this case, however, the 

 secondary minimum is better marked. The secondary minimum and maximum are distinctly marked in the 

 mean of the whole observations ; they are still existent in the mean for the 120 days least disturbed, and 

 scarcely evident in the mean of the 60 days least disturbed ; in a more careful selection, therefore, it is 

 probable that they would wholly disappear. This case is completely similar to that for the secondary 

 maximum of the magnetic declination ; and in this case, also, there is still a marked inflexion, which, however, 

 can be at once traced to the duplicate form of the diurnal variation in the winter months. The following 

 Table contains the approximate times of maximum and minimum for each month. 



TABLE XXX. — Epochs of Maxima and Minima of the Horizontal Component of Magnetic Force, 

 obtained from the Ten-Day and Five-Day series of observations, for each Month of 1844. 



From both series, the principal minimum occurs latest in the winter months, and perhaps earlier near the 

 equinoxes than at the summer solstice ; the principal maximum occurs earliest near the vpinter solstice, and 

 about two hours later near the equinoxes than at the summer solstice. 



On the whole, the principal maximum occurs earlier in the mean for the whole series, than in the means 

 for the selected series — the eft'ect of disturbances, therefore, being to accelerate the epoch of maximum. In all 

 the three series there appear to be three maxima and three minima in the month of February, and these are 

 most distinctly marked in the least disturbed, — the five-day series. These, and other more minute characteris- 

 tics, will be better considered in connection with the discussion of the succeeding year's observations. 



Intermittent disturbances seem to have less eSect on the form of the diurnal variation of the horizontal 

 component (and therefore, probably, the dip), than on the form of the variation for any of the other elements 

 discussed. 



Ranges of the Mean Diurnal Variation. — The following are the ranges of the monthly mean diurnal 

 variation, as deduced from all the hourly observations, and from the hourly observations in the ten days and 

 in the five days, selected as least affected by disturbances : — 



Jan. Feb. Marcli. April. May. June. July. Aug. Sept. Oct. Nov. Dec. 



The ranges for the months for which the sun is north of the equator, differ little from each other, and the 

 difference is least marked in the selected series. In the mean of all, January has the least range ; but the 

 means for the selected series give February the least ; in all cases, the range for June is less than for the 

 immediately preceding and succeeding months ; but this difference is least evident in the series free from 

 disturbances. 



