Ifi 



ziiiital <lis])l;v(^fmcnt of /' will introduce ;i rduplc diu' t" i^ravitv, tomliii^' to make 

 (I follow the movement of P. The peiKliiluiii swings, with the result that Q, far 

 from remainini:; at rest, may soiuetimes ae([iiire a iiHiveineiit iiiiieh t^veater than 

 that of P itself. 



We shall see later that hy eomhiniiig a eoiiinion with an inverted jiendulum 

 we may obtain a steady point in nentral eijuililninm, eapahle of l)eing used foi' 

 the measnrement of motions in any liori/ontal direction, and that the same resnlt 

 may V)e ari-ived at in other w"ays. To ohtain a steady ])oiiit with respeet to 

 niovenients in one direction oidy is, however, a sini[>ler problem, tlie solution of 

 which will l)e described first. 



§ 17. Horizoidal Voidulum Srhinuiiicfcr. 



Instead of having only one jioint fixed, let the body be pivotted in such a 

 maimer that a certain line (say tiie Htic I'Ol'm the lieure bclowl is constrain<'d 

 to remain at rest, frei'dom being left to lotate aliont this line. Further, let this 

 lixed axis be vertical and attached to the earth's suiiace: the body then forms a 

 horizontal pendulum. 



IQ 



:"s. 



IS 



II 



Then if an earthipialsc movement occurs either vertically, or horizontally 

 along (IG, the pivotted ma.ss simjiiy suffers {lis|>la<'ement as a whole. I'lUt if the 

 ground moves through any small distance horizontally and perpcndi<'idar to OG, 

 the mass I'otates through a corre.sjiondiugly small angle aliout the vcrtiial axis 



// cutting ()G produced in Q, so that Ü<1 



IS before. ^lore than this, the 



OG 



e;|iiililiriuin of the mass being neutral, no nnbalanced forces are brought into 

 action by the change of position, and there is consecjuently no reason, statii' or 

 kinetic, why the line //should move either during or after the dis])lacement of 

 the axis of su])])ort. It remains nnaffeetrd by the disturbance, and any point of 

 it mav be taken as a datum bv n'ferenee to which the motion of ncigliboiii'Ing 

 bodies lixed to the earth's surface may be determined. 



To olitain a complete nieastu'cment of anv hoii/ontal motion i.Ä the ground, 

 it is only necessary to use a ]>air of similarly pivotted and independent nia.«.se.s, 

 which may most cunvenicntly be set at right angles to each other, so that each 



