MAGNETIC DECLINATION AT BOMBAY. 
381 
Now an inspection of plate xiv. in the Philosophical Transactions for 1863 suffices 
to convince that the mean diurnal variation of declination, whilst subject at different places 
to differences of character of a minor order, is mainly of the same general type throughout 
the globe : the range may vary or even the inflections of the representative curve may be 
inverted, but the relation as to the kind of flexure of any one principal feature of the 
curve to the remaining principal features is the same everywhere : in the typical variation 
a maximum or minimum occurs between twenty and twenty-one hours, and a minimum 
or maximum between one and two hours ; and the great amount of change takes place 
between the hours seventeen and five, whilst deviations from the normal position of the 
magnets of comparatively only small extent occur during the remaining hours. To the 
same type belongs that semiannual inequality which has opposite features at the oppo- 
site solstices, and which may be called the solstitial semiannual inequality to distin- 
guish it from the equinoctial semiannual inequality, as that might be named which has 
opposite characteristics at the opposite equinoxes, — with this difference, that the turning- 
point, which in the typical mean variation occurs at twenty or twenty-one hours, takes 
place in the inequality a full hour earlier. Thus the mean diurnal variation for each of 
the half years, April to September and October to March, being the result of the 
combination of two variations of the same kind, will also be of the typical character, but 
with a small difference in the time of the occurrence of the twenty to twenty-one hours 
turning-point. Let us now examine the character of the equinoctial semiannual inequa- 
lity as exhibited in the curves 37 to 43, Plate XXX., and compare it with the type curve 
of the mean diurnal variation. In the first place, it is seen that, as in the type, there is 
scarcely any change during the night hours, and that the main variation occurs during 
half the day, in this case between eighteen hours and six hours ; secondly, the range of 
variation differs from about half a minute to nearly one minute of arc ; thirdly, the hour 
of noon is that at which the deviations due to this variation pass through zero, and on 
each side of which the inflections of the curve are inversely, but in respect to north- 
latitude stations symmetrically disposed ; fourthly, the turning-points are twenty-one 
hours and three hours, the former being a maximum and the latter a minimum for north- 
latitude stations from January to June, and for south-latitude stations from July to 
December, and vice versd, for north -latitude stations from July to December, and for 
south-latitude stations from January to June. The concurrent testimony of the five 
widely distributed north-latitude stations leaves, I think, very little room for doubt that 
this inequality is an expression of real phenomena common to the whole of the northern 
magnetic hemisphere. From the general agreement, qualified by the fact that the after- 
noon inflections are comparatively much subdued, of the curves of two stations of such 
comparative proximity as St. Helena and the Cape of Good Hope, it would be less safe 
to generalize as to the southern magnetic hemisphere ; but there is abundant evidence 
of the existence of a definite, and in some respects different law to that which prevails 
in the northern hemisphere to encourage further inquiry in this direction*. So far as 
* An examination of the Hobarton observations (including all disturbances) resulted in a curve corresponding 
closely with those of the north-latitude stations, but of small range. 
