FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900. 
225 
According to Table XVII., December has distinctly the smallest amplitude in the 
diurnal inequality in W, January coming next. June shows the largest range and 
A.D., July coming decidedly next so far as A.D. is concerned, though slightly inferior 
to August as regards range. April closely resembles the midsummer months so far 
as range is concerned, but fall distinctly short of them in A.D. In this element 
the most conspicuous difference of type is between the four midwinter months, when 
the principal minimum occurs between 10 p.m. and midnight, and the other eight 
months when the principal minimum occurs between 7 and 9 a.m. The maximum, or 
principal maximum, is the pre-eminent feature. Like the corresponding maximum of 
westerly declination, it occurs between 1 and 2 p.m. the whole year round. 
Comparing Tables XVI. and XVII., we see that the range of the mean diurnal 
inequality for the year in N is roughly three-fourths that in W ; in winter and 
equinox, however, the fraction is distinctly larger. There is less difference between 
the values of A.D. in the two elements, especially in equinox. 
§ 12. Figs. 8, 9 and 10 give the vector diagrams for the diurnal inequality forces 
in the horizontal plane. The arms of the cross are oriented in and perpendicular to 
the geographical meridian. The vector drawn from the centre of the cross, which 
serves as origin, to the points 1 to 24, represents in magnitude and direction the value 
at the hour specified of the horizontal component of the force to which the departure 
of the magnetic field from its mean value for the day may be ascribed. The scale cf 
force is such that each arm of the cross represents 10y. The vector travels completely 
round in the general clockwise direction in the course of the 24 hours. The motion 
is always direct between 6 a.m. and 10 p.m., but in the six months, October to March, 
there is a reversal of direction during some of the night hours. Details of this 
reversal are not clearly visible on the scale adopted in fig. 8, so they are shown on a 
more open scale in fig. 9, which is limited to part of the day. 
The monthly diagrams in fig. 8 bring home to the eye the great variability in the 
type of the regular diurnal variation throughout the year. If the June and 
December diagrams were alone presented, it would be difficult to believe that they 
referred to the same physical entity. There are, however, some features common to 
all the diagrams. The time when the vector is directed due south falls in all cases 
between 10 and 11 a.m., and so is well in advance of the suns crossing of the 
meridian. Again the vector in all cases has its maximum value near 1 p.m. The 
minimum value is usually seen near 6 p.m., but from April to July it occurs shortly 
after midnight. 
The May, June, and July diagrams closely resemble one another. The August 
diagram differs from them in having a distinct bay in the afternoon hours, resembling 
in that respect the September and March diagrams. The diagrams for the midwinter 
months, November to February, have a pretty close family resemblance. If we take 
the months in their order, we can recognise a continuous development from January 
to May, and from July to December; but it is obvious that in some months, especially 
