July 12, 1894J 



NA TURE 



247 



endeavoured to remedy in the latest form of the instru- 

 ment, of which a description is given in the present 

 article.' The objects of the improvements are (i) to 

 lessen the disturbing effects of the changes of tempera- 

 ture which take place in the neighbourhood of the in- 

 strument, and (2} to enable the angular value of the 

 scale divisions to be determined with greater ease and 

 accuracy. 



The chief novelty, both in the old and new forms of 

 the bitilar pendulum, is the introduction of the double- 

 suspension mirror for the purpose of magnifying the 

 movements of the pendulum. This method was used 

 by the Messrs. Darwin, at the suggestion of Lord Kelvin, 

 in 18S1, but it was also employed ten years earlier 

 by M. Delaunay at the Paris Observatory.- In the latter 

 case, however, the pendulum had a suspending wire 30 

 metres in length, and was installed in a deep pit, in which 

 the effects of changes of temperature w-ere so great 

 that the successful working of the instrument was 

 impossible. 



The Double Suspension Mirror. — A small mirror, ^[ 

 (Fig. I), is held in a light frame, from the upper edge of 

 which there project two hooks or eyelet-holes, h h'. 

 The mirror hangs by these on a fine silver wire or a silk 

 thread, w, the ends p and p' being attached to two 

 supports very close to one another. In the figure 



this arrangement is the same as that in Mr. Darwin's 

 pendulum, and therefore, in physical investigations in 

 which the double-suspension mirror is used, the vertical 

 distance between the points of support should be small, 

 otherwise the accuracy of the results may be affected by 

 the occurrence of earth-tilts. 



If the support P were rigidly connected with the end of 

 the pointer of a delicate balance, the plane of the mirror 

 being perpendicular to that of the beam, the delicacy of 

 the balance will, as Dr. Poynting has shown, be greatly 

 increased. This is, in fact, the method used by him in his 

 recent determination of the density of the earth.' Or, if 

 the support P were attached to the bob of a pendulum, 

 the movements of the latter might be magnified many 

 hundred times, and the most minute earth-tills be ren- 

 dered visible. Here we have in principle the apparatus 

 employed by Messrs. G. H.and H. Darwin 10 investigate 

 the lunar disturbance of gravity. Their inquiry, as is 

 welUknown, did not lead to the desired result, the move- 

 mems due to the action of the moon being masked by 

 the much larger ones produced by the continual tilting of 

 the earth's surface.- 



The Bifilar Pendulum.— Ihe modified form of the 

 bifilar pendulum was designed by Mr. Horace Darwin for 



these are shown in the same horizontal line ; but this is 

 not essential. All that is necessary is that the horizontal 

 distance between them should be small. 



It is evident that the wire must always lie in a vertical 

 plane passing through the centre of gravity of the mirror 

 and its frame. If, then, while the point P' remains fixed, 

 the point P be slightly displaced at right angles to the 

 plane of the wire, it follows that the mirror must turn 

 through the same angle as the line P p'. To take a 

 numerical illustration, let p p' be one millimetre, and 

 suppose the displacement of the point P to be one- 

 thousandth of a millimetre, then the mirror will turn 

 through an angle of 3' 26 , a ray of light reflected by the 

 mirror will turn through an angle of 6' 52 , and con- 

 sequently the image of the source of light upon a scale 

 three metres distant will move through about six milli- 

 metres. 



If the distance P v' be diminished, the angle through 

 which the mirror turns for a given displacement of the 

 point P is increased, and the arrangement becomes more 

 sensitive. In the position of the supports shown in Fig. i, 

 the thickness of the wire imposes a limit to the sensitive- 

 ness that may be attained. But if one support is placed 

 above the other, the horizontal distance between them 

 may be made as small as desired. It will be seen that 



^ I am indebted to Mr. IXirwin for notes about the changes made in his 

 pendulum, as well r.s for the loan of the eleciros of Figs 2-4. The pendu. 

 lum was made by the Cambridge Scientific Instiument Company One, of 

 the improved form, has recently been placed in the Royal Observatory on 

 Calton Hill, Edinburgh. 



•* C. Wolf, ** Sur un appareil propre k I'^tude des mouvements du sol." 

 Comptes Rendns, vol. xcvii. 1SS3, pp. 229-230. 



NO. I2S9, VOL. 50] 



©^' # 



f 



Fig. 3. 



the special purpose of observing and recording these 

 earth-tilts and pulsations, the points in which it chiefly 

 diners from that used in iSSi being that the mirror itself 

 is the bob of the pendulum, and the whole instrument is 

 much smaller. Fig. 4 is a sketch of the complete instru- 

 ment, while Figs. 2 and 3 show the manner in which the 

 mirror is suspended. 



A circular mirror, .M, about 20 mm. in diameter, hangs 

 by two small hooks on a very fine silver wire, w.' The 

 ends of this wire are fixed to two pomts, P P', in the 

 instrument, one of them very nearly vertically over the 

 other. The distance between these points is about 

 180 mm. If the instrument be lilted about a horizontal 

 line in the plane of the wire, the point P will be displaced 

 relatively to P' in the direction at right angles 10 that 

 plane, and the mirror will in consequence turn round a 

 vertical axis. At a short distance from the mirror is a 

 lamp with a translucent disc in front, and the reflection 

 of this disc in the mirror is observed by a fixed telescope. 

 The stand supporting the lamp and disc is moved along a 

 scale, and the position on the scale is read when the 



1 /•/,//. Trans. 1S91 A, pp. 572-574, 5S1-2 : "The Mean Density of the 

 Earth," pp. 7S-So, 83-S9. 



2 British Assoc. Report, 1881, pp. 93-126 ; 1882, pp 95-119- 



■> The plane of the mirror is arranged at right angles to the plane of the 

 suspendinc wire, in order that heat effects may as far as possible be elimi- 

 nated. tSee British .^ssoc. Report, 1893, pp. 300-301.) 



