418 EVENING DISCOURSES. 



FRIDAY, SEPTEMBER 12. 



The Gyroscope Corn-pass. By Sidney G. Brown, F.R.S. 



The lecturer spoke at some length of the magnetic or mariner's compass, 

 showing its construction ; on the variation ; on secular changes in the earth's 

 magnetism; on changes in the ship's magnetism during a voyage; on the weak 

 directive action of the earth which was further very greatly, reduced in an iron 

 ship; on the necessary corrections and their insufficiency; on the damping of 

 vibrations which caused the compass to go round with the ship so that the course 

 was sinuous and the length of a passage lengthened even with the most careful 

 steering. Even in a calm sea the ship often headed seven degrees on either side 

 of the true course. A magnetic compass was of no use in a submarine. He 

 showed a new phenomenon, that a magnetic compass of very slow period, sup- 

 ported on gymbals, even when the dynamic balance recommended by. Evans and 

 Smith in 1881 (Phil. Trans.) was perfect, might go greatly wrong if it had a 

 wobbling motion such as may be produced on a rolling, pitching ship, and is 

 quite common on an aeroplane. This was explainable mathematically ; what 

 good mathematicians had not yet explained but what he would vouch for as a 

 tact, was that if the compass were carried round in a horizontal circle without 

 wobbling, the same and even more curious errors took place. 



In an experiment here he showed a body, free to rotate about a vertical axis ; 

 the lecturer gave a small rotation, to the body, and asked the audience to observe 

 that it would continue to rotate during the whole time of the lecture. He said he 

 had proved that the solid friction was nearly infinitely small. It was this kind 

 of support that made it possible to get an accurate gyro-compass, and he would 

 describe it later. 



To illustrate some gyroscopic phenomena experimentally, the lecturer used a 

 case in which a 4-inch wheel, weighmg 4^ pounds, was revolving at 15,000 revo- 

 lutions per minute, the same as he used in the gyro-compasses which were exhi- 

 bited and would be described. The wheel was kept rotating at constant speed by a 

 three-phase electric motor fixed inside the case. The rotating part was therefore 

 not touched by brushes. The three-phase currents came from a small motor- 

 generator fed from ordinary electric light mains, the current being steadied by 

 a set of small accumulators. As being necessary for the simple explanation of 

 the compass, later, he made it clear that when the axis of the wheel was free 

 to turn in azimuth, a weight which tended to tilt the axis did not tilt the axis 

 but caused the axis to process in azimuth ; whereas a force tending to turn the 

 axis in azimuth did not do so but altered the tilt. If the axis of the wheel is 

 kept horizontal but is free to turn about a vertical axis, as the earth as con- 

 tinually tilting the axis of the wheel in space, this axis processes until in 

 the N.S. direction. When in this direction the rotation of the earth no longer 

 tilts the axis or causes motion in azimuth. Vibrations were well damped so that 

 this compass having been displaced several times returned quickly, to the north 

 every time. This proof of the earth's rotation was first described by Foucault 

 fifty-five years ago. It is an experiment which has never succeeded before in 

 front of an audience because, until now, there has never been a free frictionless 

 vertical support for a heavy body. The lecturer said that when this compass 

 ■Dointed east and west the directive action of the earth was so small a force as 

 to be only, the weight of one grain with a leverage of 12 inches. When the 

 compass pointed one degree from the north the directive action was only one grain 

 with a leverage of one-fifth of an inch. Left to itself in any position this heavy 

 compass, weighing altogether about 7 lb., reached the north in about one minute, 

 and seemed to be right with a probable error of one-tenth of one degree. 



Such a compass would be of service on land, but on a ship it needs to be 

 cRrried like a pendulum on gymbals, and its periodic time of vibration is greatly 

 lengthened The rotation of the earth does not now act directly on the gyro 

 wheel but through gravity, by means of the pendulous weight. 



The lecturer spent some time in giving a popular explanation of what occurs, 

 hut it cannot be given in a short abstract. If at any instant the angle of tilt 

 is T and the deflection from the north is A, the above experiment shows that 



