On Mcati Sea Level and its Fluetnations. 5 



Here 1 may inention. in parenthesis, that, though the fluctuations 

 which we wish to determine involve differences in the monthly means 

 amounting sometimes to only a few hundredths of a foot, it is by no 

 means necessary that the observations on which these are based should 

 be correct even to an inch. As a matter of fact, they are not so. If 

 we analyse the observations, we soon see that personal equation has 

 led to a large excess of readings at particular intervals, G in. being 

 the most frequent, and 3 and 9 in. coming next in order. But, what 



-■4 



Fig, 



1.— Mean Sea Level (in ft.) at Dundee (1897-1912) and 

 Aberdeen (1S62-1913): Mean Monthly Values eompared 

 with the Mean for the wliolc period. 



is more ciu'ious, it seems to me that, in such long series as these, 

 observations made to the nearest foot would do very nearly as well. 

 At Aberdeen, both high and low water vary over a range, of several 

 feet ; high water from something over 16 ft. to over 23 ft., and low 

 water in like manner, from about 7 to 14 ft., above the sill of the dock 

 gates. Now, over a large range such as this, the precise luimber of 

 inches will, in the long run, fall imder the doctrine of averages. 



