629 



suppose that at a particular hour, 11 P.M. for example, when the 

 amount of westerly deflections is very small, and of easterly very 

 great, the diminution of the aggregate values of either by mutual 

 counterbalance may be extremely small, while of equal absolute 

 amount in both. Now a very small amount deducted from the large 

 aggregate easterly value will scarcely have any effect whatsoever on 

 the ratio at that hour to its unit or mean hourly value ; whereas the 

 same small amount deducted from the far less aggregate westerly 

 value at the same hour would have a far more sensible effect upon 

 its ratio. Assuming, therefore, the probability that westerly and 

 easterly disturbing influences do sometimes coexist and neutralize 

 each other in the record, and that we may in some degree judge of 

 the respective amounts of the conflicting influences at the several 

 hours by the means above stated, we should be prepared to expect 

 that the ratios which are below unity do not represent the actual 

 variations of the disturbing influences at those hours quite so purely 

 as do the ratios which are above unity ; and that they are liable to 

 be affected, though in a very subordinate degree, by the abstraction 

 of the neutralized portion, when the aggregate values which they 

 represent are very small. 



"Without, however, resting undue weight upon this suggestion, we 

 may .safely say that the hours, when the ratios are below unity, are 

 hours of comparative tranquillity, and that their variations from hour 

 to hour are of a far less marked character than during the hours when 

 the ratios exceed unity. Thus viewed, the character of the disturb- 

 ance-diurnal variations may be conceived to have some analogy with 

 that of the phenomena of the regular solar-diurnal variation. We may 

 imagine the disturbance-variation (either the westerly or the easterly, 

 it is indifferent which is taken), divided as it is into two portions, by 

 the ratios being in the one case above, and in the other below unity, 

 to correspond in one of its divisions to the hours when the sun is 

 above the horizon, in the part of the hemisphere where the disturb- 

 ance may be imagined to originate, whilst the other division, or that 

 in which the ratios are below unity, and manifest hours of compara- 

 tive tranquillity, may be viewed as the hours of night at the same 

 locality. The solar hours at a station of observation which are 

 characterized by disturbance ratios above unity, will in such case 

 correspond in absolute time with the hours of the day at the sup- 



