91 



variation in the height of these datiims with respect to mean sea 

 level or to half tide level. The variations peculiar to the low and 

 high water datums, and the corrections to be made therefor, may 

 therefore be determined by taking the height of these datums with 

 respect to mean sea level, or half tide level, dm^ing the period of the 

 observations. The corrected heights, subtracted from or added to 

 the elevation of mean sea level, give them the corrected elevations of 

 the datums. 



168. Changes caused by channel improvements. — A major enlarge- 

 ment of a tidal waterway, by dredging for the improvement of its 

 navigability, may change materially the elevations of the low and 

 high water datums along it. A classic example is the effect of the 

 improvement of the Clyde in Scotland, which during the last century 

 was converted from a shallow stream, fordable at low tide, to a water- 

 way for deep-draft vessels. The enlargement lowered the low water 

 levels at Glasgow by more than 8 feet, raised the high water levels by 

 some 2 feet, and consequently lowered the midtide level by 3 feet. 

 In exceptional cases, the low water datum may even be raised by 

 channel enlargement. Extensive channel improvements may there- 

 fore require a revision of established tidal datum planes. 



MEAN LOW AND MEAN HIGH WATER 



169. Definition. — Mean low water is, as its name implies, the 

 average height of ail low waters over a long period of time, and mean 

 high water is the average height of all high waters. Because of 

 variations in the heights of high and low waters between springs and 

 neaps in the half synodic month, between perigean and apogean tides 

 in the anomalistic month, and between tropic and equatorial tides 

 in the half tropic month, a determination of mean high or low water, 

 with respect to mean sea level, from observations extending over a 

 day or a week, might differ quite widely from the long time mean. 

 To eliminate these variations the tides must be averaged over a 

 period in which these variations go through almost, if not quite, their 

 entire range one or more times. For a determination of mean low 

 or mean high water, the shortest period suitable for this purpose is 

 29 days. This period, which is sometimes called a lunation, is so 

 close to the synodic month of 29 K days as to practically eliminate the 

 spring and neap variations. It is sufficiently close to the anomalistic 

 month of 27 days, 13 hours, to nearly eliminate the perigean variation, 

 and to the tropical month of 27 days, S hours, to nearly eliminate the 

 declinational variation. Longer periods theoretically should be multi- 

 ples of 29 days. For convenience of computation it is more usual to 

 take successive 29-day periods, beginning say on the first of each 



192750—40 7 



