FROM ORDINARY DAYS OF THE ELEVEN YEARS 1890 TO 1900. 
221 
Also the diurnal inequalities in V and H follow a fairly similar course, thus we know 
a priori that the diurnal inequality in T must in general resemble somewhat closely 
that in V. In winter, however, the amplitude of the V inequality suffers a con¬ 
siderably larger relative reduction than that of H, and the difference' of type between 
the inequalities in the two elements is then at its maximum. Thus we may expect 
the resemblance of the T to the V inequality to be less close in winter, and the 
influence of the H inequality to be then more apparent. The accuracy of these 
anticipations is readily recognised if we compare Table XIII. and fig. 5 with Table IX. 
and fig. 3. 
Owing to the difference in type between the Y and H inequalities in December and 
January, their contributions to the T inequality tend sensibly to neutralise one 
another. Thus the amplitude of the T inequality at midwinter is markedly less than 
that of H, while in June and July the amplitudes in the two elements are very 
similar. Consequently the variation in the range in the course of the year is 
decidedly more conspicuous in Table XIII. than in Tables Y. or IX. 
A conspicuous feature in Table XIII. is the uniformity in the time of occurrence of 
the principal minimum. It is shown at 11 a.m. in every month except December, 
when the 11 o’clock value just exceeds that for noon. The existence of a second 
minimum in the early morning is recognisable in most of the monthly curves of fig. 5 ; 
there is at least an arrest in the rate of change of the element. The maximum, 
or principal maximum, always occurs in the afternoon, usually from 6 to 7 p.m. 
Amongst the curves of fig. 7 are three contrasting the mean diurnal inequality in T 
for the year in the years of sunspot maximum and minimum and the average year. 
The variation of type with sunspot frequency makes little appeal to the eye. 
§ 10. From the second of formulae (2) or (3) connecting the I inequality with those 
for Y and H, it appears that the contributions from Y and H oppose one another, 
but that the H contribution must be largely dominant. The phases in the I and H 
inequalities are opposite, so a comparison of fig. 6, showing the inequalities in I, with 
tier. 1 does not at first sight show the dominance of H. A little consideration will, 
however, make this clear, if we remember the difference of phase. Answering to 
the prominent, forenoon minimum in H, we have a prominent forenoon maximum 
in I, its time of occurrence being visibly earlier in summer than in winter. Again, 
as in IT, there is a distinct difference of type between summer and winter. In 
summer, the principal or only minimum occurs in the afternoon, just as the principal 
maximum does in H ; whereas in winter the principal minimum, like the principal 
maximum in H, is found in the morning. 
In most months between 2 and 6 p.m. there is at least an arrest in the fall of I, 
just as there was an arrest in the rise of H. The opposition in phase between 
the I and T inequalities is readily recognised in fig. 7, where the curves from the 
two elements representing the mean diurnal variation for the year in thea"\eiage 
year and in years of sunspot maximum and minimum are juxtaposed. 
2 H 2 
