120 MR. AIRY ON THE LAWS OF THE TIDES 



Upon inspecting the numbers in the column headed Cj, it will be perfectly evi- 

 dent that there is an error on August 9. The semidiurnal tide on that day is (com- 

 paratively) large, its whole range exceeding three feet ; and there is no instance 

 throughout all the observations of an irregularity equal to that which corresponds 

 to an anomaly of 30° with that range. I have no doubt that all tlie observations are 

 recorded too early by one hour, an error which would easily be committed at begin- 

 ning, and which, where the observations are entered on forms ready prepared, would 

 be retained to the end. Correcting on this supposition, the three successive times of 

 high water would be August 9*^ 10^^ 34'", August 9^ 22^' 50'", and August 10^ 1 1^ IS-" ; 

 and the values of q, Cg, C3, C4, for August 9'^ 22i» 50'" would be nearly 84°, 165°, 218°, 

 and 106°. 



Next I would remark that there is undoubtedly an error of the same kind in the 

 tide of July 4-5 ; but as the tide is then very small, I have not ventured to state pre- 

 cisely the alteration which I would propose. 



Thirdly, observing that where the tide is very small, the hours on successive days 

 occur earlier, in each of the instances where the order is well-marked (as from June 27 

 to July 3, from July 27 to July 31, and from August 12 to August 14), there can be no 

 doubt that the same thing must hold during the interruption of observations about 

 July 13 and 14 ; and thus it will be seen that there are certainly four high waters 

 lost at that time. I have inserted four numbers by simple interpolation, to show, 

 within two or three hours, the times of the lost high waters. 



It is also to be remembered that nothing can be inferred from such tides as those 

 of July 15-16 and July 18, where the coefficients are 0*04 and 001. 



Bearing these remarks in mind, and giving particular attention to the second half 

 of the observations, which, both for the regularity of the system pursued by the ob- 

 servers and for the agreement of the results, is greatly superior to the first, we arrive 

 at the following conclusions : — ■ 



The angle c^ increases continually, and its increase amounts to 360° in about four- 

 teen days. When its value is not far from 360°, its increase is extremely rapid. One 

 of these jumps occurs between August 2^ 9'' and August 2'^ 19'', and one between 

 August 16^ 22^^ and August 17'^ 6^' ; one also between July 2*^ 9'' and July 2^ \6^ ; 

 another takes place between July 17*^ 22^' and July 19^1 5^ but (the results being at 

 that time somewhat irregular) the time cannot be precisely pointed out. It is evi- 

 dent from this that the Courtown tides are more numerous than the Ardglass tides 

 by one tide in fourteen days nearly. 



The mean solar time of high water does not increase constantly as at other stations, 

 but oscillates backwards and forwards. Thus, from July 19 to August 18 (in which 

 period the tides at other stations have gradually retarded by twenty-four hours), the 

 evening tides always occur between 5'' 18'" and 13^' 15"^, and the morning tides 

 always occur betv/een 18'' 27'" and 24^' 14'", each time having twice oscillated be-' 

 tween its extreme limits in that period. 



