r (K^cosip, + Pcos^cos 30 «fl 
[_ -\~ (^i s ^ n tyi — P 5iW tyj) s * n 30 aj 
r (Ktdny. + PdnyJcosS 0*1 
+ "" L— C^i <** »Pi — Pcosy,) sin 30 x \ 
If we put the formula for the diurnal inequality of Table I into 
the form : 
AcosUt + Bsinlht, 
and subtract from the values A and B the annual means A' and B' 
and then again represent the differences formed in this way by: 
A - A’ = Rcos 30 ar+ Qsin 30 j 
B-B’ — R'cos 80 * + Q'sin 30 x\ 
we obtain the equations: 
• K x cos ip a -f P cos ip, = R j 
K i sin \p x — Psinip, = Q I 
K x sin tfq -f- P sin tp g = R[ i ’ 
-K.cos^A-Pcos^^Q) 
from which the four unknown quantities can be solved; then the 
amplitudes must be augmented, monthly means having been used, 
by multiplication by the factor: 
22, 
12 sin 15° 
1.0115 
and to the values ip! and tp 2 the astronomical arguments V 0 and 
V\ must be added. 
In all cases in which the S , tide is so small that even if it is 
submitted to an annual variation it can only exercise an influence 
small with respect to the amplitudes of K, and P, this simple 
method leads to good results. 
If S x is not small, we can start from the assumption that land- 
and seawind are different in winter and summer, but that they 
can be regarded as constant during each of the seasons. We can 
then eliminate out of the six summer months of the differences, (4) 
e value S x cos (15 t—C x ) and likewise out of the six winter-diffe¬ 
rences, and then we can calculate out of the equations the four 
unknown quantities. The combinations of the different monthly means 
necessary for this are of course less favourable than in the former 
case but this disadvantage can be compensated by taking a great 
num er o years together which is necessary for every method 
when we have to deal with disturbances of a meteorological nature. 
