216 POPULAR LECTURES AND ADDRESSES. 



completed through a single galvanic cell, the coil 

 of a common telegraph " detector," the metal 

 measuring-scale, the liquid, and the metal tube. 1 

 By this method it will be easy to test the position 

 of the water-level truly to the tenth of an inch. 

 It is not probable that tidal observations hitherto 

 made, whether with self-registering tide-gauges or 

 by direct observations, have had this degree of 

 accuracy ; and it is quite certain that a proper 

 method of reduction will take advantage of all 

 the accuracy of the plan now proposed. 2 



6. An observation made on this plan every three 

 hours, from day to day for a month, would pro- 

 bably suffice to give the data required for nautical 

 purposes for any harbour. It is intended immedi- 

 ately to construct an apparatus of the kind, and 

 give it a trial for a few weeks at some convenient 

 harbour, and if the plan prove to be successful and 

 convenient, it will come to be considered whether 

 observations made at every hour of the day and 

 night might not, all things considered (accuracy, 

 economy, and sufficiency for all scientific wants) 

 be preferable to a self-registering tide-gauge. 



1 Instead of the galvanic detector, a hydraulic method may be 

 found preferable in some places. The latter consists in using a stiff 

 tube of half inch diameter or so, instead of the solid metal 

 measuring-bar, and testing whether its lower end is above or below 

 the level of the water by suction at the upper end. 



2 The "Clyde Trust" have given permission to try this method 

 at a convenient place in the harbour of Glasgow. It is probable 

 that it will also be tried in the harbour of Belfast. 



