(5} 
The theoretic coefficient of importance ') expressed in the mean 
amplitude of M t as unity is 
for K x ' 0.58 
for P: 0.19 
so that yearly twice the two tides strengthen each other, namely in 
June and in December, when the relative amplitude is : 0.77, whilst 
in March and in September it comes to: 0.39. 
There must therefore be, even though the meteorological tide S 1 
were constant during the whole year, a considerable annual variation 
m the diurnal inequality. 
Although during the period under treatment the lunar declination 
was particularly small and the circumstances therefore were un¬ 
favourable for the determination of the constants of the sidereal tides 
yet the theoretical amplitude of K x , expressed in deviations of the 
vertical pendulum, amounts for Potsdam to: 
0".0050 
whilst the amplitude of P is in square numbers : 
0".0020 
The amplitude of the annual variation 
must therefore be about: 
the diurnal inequality 
0".0070 
i. e. almost twice that in the principal solar tide S 3 : 
0".0040. 
From the amplitudes of the diurnal movement of Table I is 
evident that the amplitude of the annual variation : 
June—December = - — 0".0081 
differs but little from the theoretical value, so that we have every reason 
o make an attempt for an accurate determination of the constants 
Which promises to lead, at all events for the tide 1C , to satis- 
factory results. 
• ?n e ; alUe 0f Such an invest igation is not so much to be found 
m the determination itself as in the fact that, if the investigation is 
“-V 6 - 1 >' ears 80 ,liat the irregularities of meteoro- 
gln have dlsa PP eared ’ we shal l be able to correct the 
onthly means of the diurnal inequality for the influence of the 
nu,X°,rT UdeS ’ in K 0rder t0 obtain this way accurate series of 
be smri /X 7,“‘ nature and the or, 'g in ° ( *be tide can 
be studied and deduced. 
papers, Cambridge. 1907, vol. I. p 25. 
