4l8 On tbt Mechanifm hy luh'tchtht Manner's Compafs it fufpenJcd. 



liberty to move : — that is to fay, if the weight of the box or its psrts be properly dlfpofed, 

 the compafs will aflume a pofirion in which its upper furface fliall be horizontal. 



The fame principles which were applied to the (ingle centre of the magnetic needle will ' 

 alfo apply to the axis of the gimbals. If the centre of gravity of the conipafs-box be fo 

 placed with rcfpe£l to either axis as that its vibrations (hall be quick, every horizontal action 

 will gre.Ttly difturb it, and it will not fpeedily fettle. The moll favourable pofition of the 

 pivots or edges of fupport in the gimbals will be when they all lie in the fame plane, and 

 the centre of gravity of the compafs-box is very little below that plane. 



The pra<SicaI application of thefe inferences appears to be, that, without pretending, as 

 has been done, that any peculiar fecret or great difcovery is required to give lability to this 

 ufeful inftrument, nothing more is required than good workmandiip, and a proper adjuft- 

 ment of tlie weight with regard to the centres or axes of fufpenfion. The needle ought to 

 be adjufted either by means of its cap, or by proper filing away, or elfe by additional pieces 

 to the card, fo that it fhall vibrate very little, and that flowly, when placed upon a point and 

 moved horizontally, whether in the direflion of the needle or at right angles to that direc- 

 tion. The card is then ready for the compafs-box. The box itfelf mud be adjufted witli 

 the card in its place, fo that it fliall exhibit the fame ftcadinefs when moved in the line of 

 dlreftion of the outer pivots. And laftly, the fame difpofition mud be made with regard 

 to the motion in the direflion of the inner pivots. It is fcarcely necelTary to add, that the 

 means of this adjuftment confift in fliifting the pivots tliemfelves, or, which is much better, 

 in altering the difpofition of weight about the compafs-box. An external ring of metal, 

 rticircling the box, and raifed or lowered until the proper place for fixing it is found,, 

 would perhaps afford the mod convenient method. 



,Upon the whole, the reader will perceive, that the leading aim of the prefent paper is to 

 enforce the truth, that the compafs is very little didurbed, at fea or elfewhere, by tilting the 

 box on one Cde, but very Ttinch by fudden horizontal changes of place; and, confequently, that 

 a fcientific provifion againd the latter is the chief requifite in a well-made indrument of this 

 kind. And again, that nothing is more eafy than to afcertain whether a compafs polTefles 

 ftability ; fince nothing more is requifite than to Aide it upon a table in the fcveral drreflions 

 above-mentioned, and remark how far it is didurbed. The good workmandiip of the cap ■ 

 and pin of the needle may be afcertained by infpeftion with a magnifier, and alfo by 

 dlrawing the card with afmall key or other piece of iron, a very little quantity, for example 

 a quarter of a degree, out of its ftation or pofition, and remarking whether it returns acca^ 

 rately to its original datton. 



Before I quit this fubjeft, I mud take notice, that, as the fufpenfion on a point has been ■ 

 applied to the compafs-box, fo, on the other hand, the gimbals have been applied to the nee- 

 dle. This was done by tlie late Dr. John Lorimer, who difpofed a dipping needle on its 

 own axis between the cheeks of a frame parallel to the diameter of a circle of brafs gra- 

 duated to (hew the angle of dip. This circle or meridian had pivots at its zenith and nadir, 

 which moved in holes diametrically perforated in another circle. The needle was by this 

 means not only permitted to difpofe itfelf in a line forming the angle of dip with the hori- 

 zon, but was enabled to carry the meridional circle into the line of the magnetic meridian. 

 It does not appear from the Do£lot's paper and drawing in the 65 th volume of the Philofo- 



pbiicat 



