rentn' i>f wliidi a brass tube d (titf. "Jo I is fixwl. The uj)]>er end of tlie multiplv- 

 iujl lever is a brass ball fitting easily l>ut not loosely into this tube: it stands at 

 sufh a height iu the tube as to be at the centre of percussion i){ tiic jiendulura. 

 The fulcrum of the multiplying lever is a gimbal joint isee fig. 2()1 giving freedom 

 of rotation in any a/imnth, i)Ut no freedom to rotate about a vertical axis. This 

 is carried by a stitl' iron bar c projecting from a wooden ui)right /, which is 

 lirndy luiih-d to a jxist </ (fig. 241 stuck fast in the ground. The projecting bar 

 <" is slotted where it is fixe<l to /' l)v a sercw-bolt, so that the j)ositioii of tlu' 

 gimbal joint may be adjusted horizontallv, bv turning ./ round or i)Ushiiig it out 

 and in. I'clow the fulcrum the niuiti]>lying l<-ver consists of a liglu bamboo rod 

 h, terminated bv a iork in wiiicji the marking intlex i is jointed on a horizontal 

 axis. Th<' jmintci' is a piece of .straw tipped with -ted : it rests on the smoked- 

 gla.s-; plate J, and its ])re.«sure is adjustable by a small counterpoise /,-. The 

 glass ])latc is supported in the manner already shown in Plate 11, and stands on 

 a boani fix<'d to the top of the |>ost ij. It receives continuous motion from a 

 clock, by means of a müer, in precisely the same manner as the glass jdate in 

 Plate II. The plate i.-. kept continuously revolving, in expectation of an earth- 

 ipiake. and it is fi>ui\d that altliough the whole apparatus is remarkably free 

 from friction, the line traee<l on the plate does not widen to any very objec- 

 tionable extent, ill the al)sence ot carth(|iiakes. The multiplying ratio is ten 

 to one. 



The additional, or alternative, ai-rangement shown in tigs. 27 and '2H, 

 allows the .sinie |iendii]um to mm-vi' as a two-componciit machine. A brass plate 

 / is fixi-d to th<' cro.ss-bar •■ at the level ol' the c<'ntre of percu.ssion of the pendu- 

 lum. In tills two slots are cut, at right angles to each other (see fig. US, where 

 a [xirtion of the b;ir c is remnvedl, and in the-e --lots tln' bentup ends of two 

 short horizontal levers m, n, slide, w and /( .ire pivotted about vcitical a.\es at 

 ri and p. The supports of the.se a.vs are fi.xid to the earth. Heyond the axes o 

 and p, to the right, are two light continuations of the levers in ami ji, arranged 

 IHirallel to iwh other. The.s«' iq and i\ record their motions on a large revolving 

 .-inoked -glass jilate. a portion of whicii is shown at .^■. It will be obvious that a 

 horiz/mtal motion of the earth parallel to » will simplv cause ii to slide in the 

 plat«- / without alli-cting the pointer ;■, but will cause )ii to revolve about th<' 

 axis o, and will therefore l)e recorded (.n a magnified scale by tln' poimci- y. 

 One of the two levers m >i is bent so a.s to |)a.ss bi'low the other without toui'hing 

 it. The upright ends by which they receive the motions of the bob are cylinders, 

 working ex-ily. but without shaking, in the slots in the plate /. The long 

 |M>intcrs )• and q are jointed horizontally near n and ;), (see q. (ig. '_*H) .-o that 

 they may follow ineijiiulities of level in the revolving gla.ss plate. 



The inertia of the ImiIi i- is so great that there is no objection, on the .-core 

 of friction, to the nimultaneoii.s us«- of lM)th arriingc-nient.s for recording, and they 

 are i-on.KtriK-ted so a.« not to interfere with each other. The recording l(-v(-rs are 

 .'^> light that their intliienie. ImjiIi .-tutic and kinetic, on the bob is negligible. 



