( 464 ) 
same direction during a certain number of years; thus four successive 
years often occur with high or low level; e. g. 1857—60,1866—69 
1877—80 and 1903—06. 
A comparison of some years, e. g. of 5 years with some other 5 years 
can consequently give rise to great errors; supposing one period of5 
years is on an average 3 cm. too, high through the influence of the 
wind, the other period of 5 years on an average 3 cm. too low through 
this influence, this gives a difference of 6 cm.; if now the sinking of the 
ground during that time is 6 cm., we shall find the sinking of the 
ground to be =0 if the high years are the first, but =12 cm. if 
the low years are the first. It is therefore desirable to choose the 
longest periods possible for a comparison, for the periods in which 
deviations in consequence of the wind take place in positive or 
negative sense, may be of pretty long duration. A fixed rule for 
these deviations or fixed periods for them could not be found. Hence 
the best thing is to compare the mean of a certain number of years 
with the mean of another number of years. 
A proof that the deviations in consequence of the wind can be in 
the same sense during many years is furnished in the difference of 
the mean monthly height of the months inter se, in which all the 
errors by displacement of the zero point are eliminated. Thus I have 
found for a 54-year period 1843—1896 of the observations on the 
gauge at Cuxhaven, that in the first 27 years the mean position in 
February was 154 mm. higher than the mean position in May, in 
the last 27 years 43 mm. lower, thus presenting ,a difference of 
197 mm. 
In order to derive the probable sinking of the ground from the 
Observation on different gauges, we can, in order to eliminate as 
much as possible the errors caused by the shifting of the zero points, 
take the mean of the observations of a certain number of gauges, 
or draw a conclusion from every gauge and take a mean of these 
conclusions. For a good survey and comparison, as the deviation in 
consequence of the wind is not everywhere the same on account of 
the local conditions, and to enable us to use the observations on 
that gauge as much as possible for which the longest series is avail¬ 
able, the last way is preferable. Of course no gauges are to W used 
at places where the mean level is modified for some reason or other, . 
as e. g. at Zoutkamp. 
When we make a graphical representation of the observations on 
different gauges, it appears that of late years the sea-level has very 
decidedly risen with regard to the land. The extent of this rise cannot 
