CHAP. IX.] TIDES 233 



mean water-level, as found by meaning the height of one high 

 water with the following low water, will differ from that found 

 by meaning the same low water with the following high water. 



A fair approximation to the true mean water-level may be 

 obtained by taking the mean of the results for successive tides 

 obtained as above for a period extending over a considerable 

 interval, especially if the barometric height be allowed for ; 

 the longer the period the better. 



When mean water-level has been found, it should be referred 

 to the Ordnance Datum in the United Kingdom, or abroad to 

 some fixed plane of reference ; and for this purpose natural 

 marks should be preferred, such as rocks which cover or un- 

 cover at certain conditions of tide. If no natural marks are 

 available, artificial marks should be created. The mark 

 adopted should be stated in the memoir of the chart. 



In some cases the mean level of the water may be made use Mean 

 of as a temporary datum for reducing the soundings. D^tum ^ 



If, for instance, we commence soundings in a place where 

 we do not yet know the spring's range, bu*. intend to get it 

 accurately after some months' observations, we may find it 

 convenient to reduce all soundings to the mean level, as 

 found by meaning each day's high and low water. Then, 

 when we have ascertained the level of low springs below this 

 mean level, one uniform quantity will have to be subtracted 

 from every sounding, which will save a good deal of complica- 

 tion and waiting, as the soundings may all be plotted without 

 fear of mistakes in reducing them afterwards. 



This mode will be mostly used for shallow channels, where 

 a difference of a foot or two is an important matter, but it is 

 liable to the error caused by variation in the height of the 

 mean tide-level. 



When a long series of observations is plotted on the abstract Analysing 

 form provided for the purpose, any peculiarity in the day and Sjat^^ 

 night tides should be noted in connection with the sun and 

 moon's declination. 



Other peculiarities connected with the moon's declination 

 should also be noted if they exist, such as that of the diurnal 

 inequality of height of low water being greater than the diurnal 

 inequality of high water, or cases in which the diurnal in- 

 equality in the times becomes so large that there is only one 



