204 



NATURE 



\_Dec. 27, 1883 



the third the observed motion per hour, and the fourth the calcu- 

 1 ated motion. The table has been so drawn up that it begins with 

 l>laces nearest the earth's equator and passes gradually to others 

 further away, going from Ceylon at b" N. lat. to New York at 

 40° N. lat.. New Haven at 41°, and ending with Aberdeen at 

 57°. At the first-named place it will be seen that the pendulum 

 .swings through less than 2° per hour, whilst at Aberdeen it swings 

 through nearly 13°, which is an approximation, at least, to the 

 statement I have made, that, since the rotation of the pendulum 

 plane will be most rapid at either pole, the further from the 

 equator we swing it the greater will be the number of degrees 

 passed over per hour. 



To turn now to the gyroscope. We shall expect, if we succeed 

 in imparting to it a rotation which is independent of and un- 

 affected by the earth's rotation, that the angular change shown 

 by it will be the same as that indicated by the pendulum, or, in 

 other words, that the number of degrees passed over will be the 

 same in both cases. 



In the gyroscope, that portion which corresponds to the 

 swinging part of the pendulum i^ the heavy disk seen in Fig. 28, 

 to which a very rapid rotation can be imparted. This disk is 

 mounted upon the horizontal circle shown in the figure, which 

 circle in its turn is mounted in a vertical one suspended by a 

 bundle of raw silk fibres which depend from the little screw 

 shown at the top. by means of which the vhole system can be 

 raised, so preventing the vertical circle from resting its whole 

 weight upon the piivot below, the use of which is not so much to 

 support the apparatus as to guide it in its movements. 



Now ij order that the rotation of the disk shall be uninfluenced 

 by the motion of the earth a great number of precautions have 

 to be taken. The first of these is to insure that the whole of 

 the appaiatus shall be perfectly free to rotate, and that, however 



e> epiece. 



much the silk fibres supportirgthe vertical circle may be screwed 

 up in order that it may not rest its weight upon the pivot, its 

 molion shall not be interfered with — that there shall be no twist 

 in the thread. This is the first precaution ; and, when this has 

 been done, a condition of things is obtained in which the apparatus 

 is perfectly free to move round a vertical axis represented by the 

 .silk fibres prolonged. Then, having fulfilled this condition, the 

 next matter of importance is to see that the disk is perfectly free 

 to move on the horizontal axi-. For this purpose the wheel 

 which holds the two extremities of the axis of the rotating disk 

 is armed with counterpoise weights (see Fig. 28), two in a 

 horizontal plane, A A, and two in a vertical plane, of which one 

 is seen at B. 



Then the knife edges, c c, which are exactly in the plane 

 of the centre of motion of the whole system, are made to rest 

 on two steel plates mounted on a separate stand, in order 

 to ascertain if the moving parts are perfectly balanced, the 

 perfection of balance being determined by the slowness with 

 which it oscillates up and down. But this is not all ; 

 it must not only fe so adjusted by these weights, A A, that 

 the ring shall remain horizontal, but it must be so perfectly 

 balanced by the two weights, one of \\hich is seen at B in 

 Fig. 28, that if a considerable inclination be made from the 

 horizontal it will be taken up equally on Ijoth sides. Finally, 

 the instrument must be so adjusted that when the two delicate 

 knife edges are placed on the two steel plates in the outer ring 

 (see Fig. 28) the ring carrying the disk shall be perfectly free 

 to move and have its centre of molion exactly identical with the 

 centre of motion of the outer ring and of the disk itself. Then, 

 "hen all these precautions have been taken, and the disk is set 

 rotating with considerable velocity by means of a multiplying 

 w heelwork train, we have, as far as the mechanics of the thing are 

 concerned, an experiment juat like the other, with this import- 

 ant diffeience, however, that, whereas the pendulum experiment 



always succeeds, much trouble is often experienced in experi- 

 menting with the gyroscope. But, when the multiplicity of the 

 conditions necessary to the success of the experiment is considered, 

 this is not surprising. If, however, all the conditions have been 

 adhered to, the pointer with which the instrument is fitted 

 (see Fig. 29) ought to niove over the scale at exactly the same 

 rate that the pendulum moves over the scale beneath it. But 

 even supposing that the pointer of the gyroscope does move over 

 the paper and in the right direction when the apparatus rotates 

 one way, this is not enough. The demonstration of the validity 

 of the result given by it is that an equivalent deviation is obt.iined 

 when the apparatus is turned about in every possible direction. 

 The first test of course is to rotate in the opposite way, then, if 

 all the adjustments have been properly made, the deviation ob- 

 tained will be the same in amount and direction as before, and it 

 may be taken that the result obtained is then really due to the 

 earth's rotation. 



With this reference to the most important points connected 

 with the gyroscope, we may bring our inquiries under this head 

 to a close. So many men have worked with the instrument in 

 so many lands, and under such rigid conditions, that there can 

 be no doubt that the rotation of the earth is demonstrab'e by it, 

 although certainly its verdict is not anything like so sharp, or so 

 clear, or so ea-ily obtained, as that given by the pendulum. 



Our appeal to jihysics has at once put out of court the old 

 view of the arrangement of the universe, which placed an im- 

 movable earth at its centre. How Copernicus w as the first to 

 point out that this old view was incorrect, and that it was the 

 earth \\ liich moved, and how Galileo was persecuted because he, 

 in times much less for tuna' e than our own, had the courage to 

 say so, — these are familiar points in the history of the discovery 

 of the earth's rotation. 



Having then demonstrated the existence of this particular 

 movement of the earth, we must now proceed to a consideration 

 of the rate, directim, and results of the movement, — connect 

 in fact the pendulum of Foucault with that of Huyghens, and 

 regard the physical pendulum as giving an important use to the 

 experiments of Galileo and of Huyghens in which they caused it 

 to act as a controller of time. 



Turn back to our two tables. They are not without interest 

 at the present moment. In the first table, " Hourly Motion of 

 Pendulum Plane," the observed motion of the pendulum plane 

 per hour is connected with the latitude of the place at which it 

 swings, varying as that varies; and therefore the observed motion 

 ill any latitude ought to give the same value for the earth's 

 rotation, the closeness of wh'ch to the real value will at the 

 same time be a measure of the accuracy of our pendulum 

 observations. 



Let us endeavour then to find out in what time the earth must 

 go round in order that the pendulum plane may vary (say) 

 ■A° P'^'' hour in Ceylon, w^^" in Dublin, and so forth. 



Taking our clock as being divided into twelve hours, each 

 hour into sixty minues, and each of these again into sixty 

 seconds, it is found (see Table 2) that the value for Ceylon is 

 23h. 14m. 20s., and for Dublin 24h. 14111. 7s., the mean value 

 ol the observations at the various places mentioned in the 

 table being 23h. S3m., so that according to that table the 

 earth rotates on its axis in a few minutes less than twenty-four 

 hours. 



Now although such an approximation to the real value may 

 suffice for the great mass of mankind, it is not an astronomical 

 way of dealing with the question. We have seen the circum- 

 ference of a circle divided first into degrees, then into J degrees, 

 next into seconds, ai d finally into tenths of seconds ; by the 

 application of electrical principles, time has been even more 

 finely divided, and the ques'ion naturally arises. Are there any 

 means of determining the exact period of the earth's rotation ? 



There are means of doing this. In the last lecture occasion 

 was taken to point out that the stars are infinitely removed from 

 the earth ; the stars being so infinitely distant, a slight change 

 in their position will not be perceptible to an observer on the 

 earth, and the place of a star to-day and its place to morrow are 

 the same so far as relates to any parallactic change of position. 



This being premised, it w ill lie clear that, in order to get out 

 the exact period of the earth's rotation, one only has to make an 

 observ.ation of any star on one particular day (such observation 

 being of course made with a clock), and repeat the observation 

 when the star is in the same position on the succeeding day. 

 The time which elapses between the observations must be the 

 time taken by the earth to make a complete rotation. But it 



