THE TIDES. (A PP. D.} 215 



But direct observation and measurement would 

 probably be more accurate than the records of the 

 most perfect tide-gauge likely to be realized. 



5. One object proposed for the Committee is to 

 estimate the accuracy, both as to time and as to 

 scale of height, attained by the best self-registering 

 tide-gauges at present in use, and (taking into 

 account also the relative costliness of different 

 methods) to come to a resolution as to what method 

 should be recommended when new sets of observa- 

 tion are set on foot in any place. In the mean time 

 the following method of observation is recommended 

 as being more accurate and probably less expensive 

 than the plan of measurement on a stem attached 

 to a float, often hitherto followed where there is no 

 self-registering tide-gauge. A metal tube, which 

 need not be more than two or three inches in 

 diameter, is to be fixed vertically, in hydrostatic com- 

 munication by its lower end, with the sea. A metal 

 scale graduated to centimetres (or to hundredths of a 

 foot, if preferred) is to be let down by the observer 

 in the middle of the tube until it touches the liquid 

 surface ; and a fixed mark attached to the top of the 

 tube then indicates the reading which is to be taken. 

 Attached to the measuring-scale must be one or 

 more pistons fitting loosely in the tube and guiding 

 the rod so that it may remain, as nearly as may be, 

 in the centre of the tube. The observer will know 

 when its lower end is precisely at the level of the 

 surface of the liquid, by aid of an electric circuit 



