MR. LUBBOCK ON TIDE OBSERVATIONS. 277 



before high water is expected, all which is afterwards copied into a book ruled in 

 the same manner, and the time of high water, with the height, is easily inferred. 

 The height is ascertained by means of a rod or tide-gauge, connected with a float, 

 which is placed in a chamber, into which the water enters through a culvert, so that 

 the ripple or agitation of the water in the river is avoided as much as possible. A 

 clock, carefully regulated, stands close at hand. This plan has been adopted at 

 my suggestion, and, if the observer and the transcriber of the observation do their 

 duty, it does not seem to me to be susceptible of any improvement. 



I find by examining the registers of the observations at the London and the St. Ka- 

 therine Docks, that the tide is on the average about five minutes later there than at 

 the former place, and the difference in height between the lines or zero points, from 

 which the rise is measured, is about five feet. Hence I formed Tables A. and B. by 

 first adding five minutes to the time of high water, and five feet to the height of high 

 water given in the London Dock books, and then comparing the times so altered 

 with those of the St. Katherine Docks. The discrepancies exhibited by these Tables 

 may be attributed to the carelessness of the observers at one or both places, or to 

 the inevitable difficulty of measuring with precision the quantities sought. I next 

 formed Tables C. and D. by adding five minutes to the time of high water given in 

 the St. Katherine Docks register, and five feet to the predicted heights in the British 

 Almanac, in order to compare the observations with the predictions given in the 

 British Almanac, and which are founded upon the Tables, the construction of which 

 I have explained in the Philosophical Transactions, excluding only those observations 

 which, by discordance with the cotemporary observations at the London Docks and 

 with the prediction, appear doubtful. It results from this comparison, that the average 

 error of the predictions of the time of high water contained in the British Almanac 

 is about ten minutes, that is, when the plus and minus errors are not allowed to 

 counteract each other. The average error of Mr. Bulpit's Table is about the same. 

 The tide predictions of Mr. Epps are evidently founded upon the same Tables and the 

 same methods with those in White's Ephemeris, and do not agree quite so well with 

 observation. Those of Mr. Innis are more inaccurate. 



When the plus and minus errors are allowed to counteract each other, the average 

 error of the predictions in the British Almanac being, from the observations of the first 

 six months of this year, —9 minutes, leads me to suspect that a change in the establish- 

 ment has taken place owing to the removal of the old London Bridge. If I am right 

 in this conjecture, it is worthy of remark how sensibly the phenomena of the tides 

 are affected by any slight alteration of local circumstances. Moreover, the height of 

 high water appears to be less by 2 inches than formerly. If the predicted times be 

 increased by 9 minutes and the heights be diminished by 2 inches, the predictions 

 will then agree with observation, nearly as well as the observations at the London 

 and St. Katherine Docks agree with each other. 



