10 UNITED STATES COAST AND GEODETIC SURVEY. 



sliding frames, pulleys, and a summation chain. This system is 

 operated simultaneously and with the same hand crank as for the 

 first side, and indicates by means of an electric circuit and electro- 

 magnets the times of the high and low waters. At these times the 

 machine is automatically stopped by a brake, which prevents the 

 operator from thoughtlessly running by these points. 



To set up the machine with the harmonic constants, predict, and 

 tabulate the high and low waters for a year at any station requires 

 from 10 to 15 hours, depending upon the complexity of the tides at 



the station. . 



PREVIOUS TIDE-PREDICTING MACHINES. 



The machine to be described here, like almost every contrivance, 

 apparatus, or machine in practical use, is based very largely upon 

 what has been accomplished by othei-s who previously labored in 

 the same field. 



When harmonic analysis had been applied to the reduction of 

 tidal observations, and it became practicable to take into account 

 more than the most important elements entering into the composi- 

 tion of the tides, the need of mechanical means for relieving the 

 brain of the methodical and wearying labor involved in their pre- 

 dictions became at once apparent. 



BRITISH TIDE PREDICTOR NO. 1. 



In 1872 Sir William Thomson (later Lord Kelvin) had made a 

 rough model, with 8 components, for showing the feasibility of pre- 

 dicting tides mechanically. Under the auspices of the British 

 Association he designed and had made in 1873 by A. Leg6 & Co., of 

 London, the first tide-predicting machine, containing the 10 com- 

 ponents M2, S2, N2, K2, Qi, Ki, L2, Pi, M'i, and (MS)^. He gives 

 credit to ^'A description of a machine for finding the numerical roots 

 of equations, etc.,'^ by the Rev. F. Bashford, read before the British 

 association in 1845, for suggesting to him the means of summing 

 mechanically the series expressing the tidal fluctuations of the sea 

 employed in his machine. It is known as the British Association 

 tide-predicting machine or tide predictor No. 1. (See plate 1.) It 

 is described in Thomson and Tait's Natural Philosophy, second 

 edition, Part I, pages 479, 480. Having been used but little, it was 

 deposited in the vSouth Kensington Museum. 



BRITISH TIDE PREDICTOR NO. 2. 



In 1879 a second tide-predicting machine was designed by E. 

 Roberts, of the British Almanac office, and constructed under his 

 direction for the Government of India by A. Lege & Co. It had 

 originally 20 components, but in 1891 the component Xj was replaced 

 by T2, and 2N, MN, 2MK, and MK were added, making in aU 



