2 FISCHER: TIDE PREDICTING MACHINE 



1913. It is now dismounted and beingput in a finished state for 

 final installing in the early part of next year. 



Six other machines for predicting tides now exist. The first 

 one, designed and constructed in 1876, by Sir Wm. Thomson, 

 later Lord Kelvin, containing 10 components, was but little used 

 and is now on exhibition in the South Kensington Museum, Lon- 

 don. A second one, wit^i 20 components, was constructed in 

 London for the British India Survey in 1879. The number of 

 the components was later increased to 24. It has since been in 

 use for predicting the tides of the British India ports. The third 

 machine was constructed in 1880, by Sir Wm. Thomson. It 

 contains 16 components, and it is supposed to be in use for the 

 tides of the British ports. A copy of this machine was acquired 

 by the French government in 1900 for predicting the French 

 tide tables, and another of the same style was made for the gov- 

 ernment of Brazil about three years ago. 



All these machines produce automatically a tide curve, which 

 must be scaled off for the heights of the high and low waters, 

 their times being taken, in the simpler tides, from hour marks upon 

 the curve. For exact times a second curve, that representing 

 the first derivative of the height series, must be traced and scaled 

 off. While it takes only from two to three hours to produce a 

 year's tide curve, several days' work is required to prepare the 

 printer's copy from that curve. 



To obtain data required for the printer from the machine direct 

 led Prof. Wm. Ferrel of the Coast and Geodetic Survey, when 

 the need of a tide predictor in that Bureau became imperative, 

 to design one upon principles differing considerably from those 

 of the English machines. In it the sine terms of the derived or 

 time series are summed simultaneously with those of the cosine 

 or height series, the face of the machine indicating the exact 

 times of high or low water to be copied by the operator upon the 

 form for the printer. For the simpler tides the heights are also 

 taken from the face by the use of an auxiliary scale, but for the 

 more complicated tides, the machine requires a second setting 

 for the exact heights. This machine was constructed in 1882 

 and was used for preparing the tide tables of the Coast and Geo- 



