Magnetic Instruments on the Galilee 25 



The foregoing description will serve to set forth the main features and fundamental 

 principles of the new device. With the steady improvements subsequently^ made, it was 

 found capable of yielding a satisfactory degree of accuracy. The liquid of the compass acted 

 as an excellent damping device, so that settings could be made with great rapidity and ease, 

 about ten minutes sufficing to give an approxunate value of H. The form of mounting 

 the deflecting magnet so that it partook of the motions of the compass bowl not only tended 

 to preserve the requisite constancy in the distance between the centers of the card and 

 magnet, but also largely avoided the troublesome effects of the rocking of the deflected 

 magnet during the ship's motions, which occurred in the style of ship magnetometer where 

 the deflecting magnet was mounted off to one side and not constrained to move with the 

 bowl. 



The deflecting magnet for instrument Dl was mounted in a paraffined wooden block, 

 to the bottom of which was fastened a disk, with lugs which fitted in holes in the disk on top 

 of the bridge or supports, so as to admit of putting block with magnet in an invariable posi- 

 tion, direct or reversed. The thermometer was inserted in a hole in this block (see Plate 5, 

 Fig. 1). It will be observed that the magnet, when placed inside the block, did not require 

 thereafter to be touched during a complete set, the entire block, with magnet inside, being 

 reversed. 



The method employed for measuring H at sea is not an absolute one, but depends 

 upon a knowledge of the magnetic moment of the deflecting magnet and its variation with 

 time; hence various styles of magnets were employed, and also two deflecting distances were 

 provided, a separate block for each magnet having been constructed. At first, four magnets 

 were used: a hollow, cylindrical magnet,' made by Tesdorpf as an auxiliary magnet for one 

 of his field magnetometers, this magnet being designated No. 45; next, a long-tube as well 

 as a short-tube magnet, each consisting of a bundle of magnetized wires, such as are used 

 in liquid compasses; and, finally, intensity needle No. 4, used with the sea dip-circle. The 

 latter needle was used because its magnetic moment could be completely controlled by the 

 loaded-dip and deflection observations with the sea dip-circle, a combination observation 

 of dip circle and deflector thus practically amounting to an absolute method. 



Extensive experiments were made with the four magnets at Washington and at the 

 Cheltenham Magnetic Observatory in the spring of 1905; these experiments were repeated 

 on Goat Island, San Francisco Bay, in July 1905, after the instrument had been brought 

 across the continent. The results were so satisfactory that it was possible to restrict the 

 ship observations to magnet 45 and the long-tube magnet NL. In spite of the obvious advan- 

 tage above noted, it was finally deemed best not to use intensity needle 4, because the dip- 

 circle needles should not be handled any more frequently than necessary, in order to safe- 

 guard against injury to the exceedingly delicate pivots and the consequent vitiation of a 

 whole series of observations. 



At every port corresponding observations were made ashore with the sea deflector and 

 a standard land magnetometer, and thus a combined intensity-constant was determined at 

 various temperatures, this constant being a function of the magnetic moment of the magnet, 

 deflection distance, etc. The experience as to constancy of magnetic moment with magnet 

 45 was especially satisfactory. The complete set of observations also called for rotation of 

 magnet within its block through 180°; in other words, every means was taken to eliminate 

 possible changes. 



One other point requires mention. On board ship, the line of reference — the lubber- 

 Une — is turning with the ship, so that the deflection-angle deduced from two consecutive 

 settings, e. g., a and 6, would require a correction equal to the difference between the ship's 

 headings at the two settings. The possible errors in the deflection-angle, caused by changes 



'This magnet was courteously supplied by the United States Coast and Geodetic Survey. 



