4.26 On the Mechanifm iy which the Mariner's Compnfs is fufpeudcd. 



column of water equal to one of the lifts, and that the quantity of fteam to be difcharged 

 at caih Ilroke muft occupy a fpace equal to that of all the water moveil at each (troke, and 

 muft in all cafes be conliderably llronger than the atmolphere. 



VII. 

 On the Mechanifm by which the Manner's Compafs is ftifpendcd. 



Xf a bar of iron be rendered magnetical, by rubbing it on a natural magnet, or by any of 

 the well-known procclTes for that purpofe ; it acquires the property of difpofing itfolf nearly 

 in a north and fouth line, whenever it is fufpcnded fo much at liberty that the energy of 

 this power is fuHicient to overcome the friclion or other impediments to its motion. The 

 mariner's compafs is an apparatus in which a bar of this kind, called the needle, is fup- 

 portcd, for the highly ufeful purpofe of determining the pofition of the meridian at fea, and 

 confcquently of enabling fhips to fteer their courfe by day or night, without obfervation 

 of the ftars, or any other external obje£ls, as was neceflary before the difcovcry of this in- 

 ftrument. In a well-con(lru£led mariner's compafs the needle is defended from the im- 

 pulfe of the air, and is little fubjefl: to be difturbed by the external motions or agitation of 

 a fliip at fea. As this difturbanee is, however, the chief impediment to the convenient 

 ufc of the compafs in a boat, where the motions are fudden and fiiort, or in a fliip,when the 

 ■waves are very turbulent ; and as the artifts in this branch have endeavoured to perfuade 

 the world that certain pieces of mechanifm were much fuperior in their ufe to others dif- 

 ferently difpofed ; — I thought it might be of fome utility to explain the fimplc principles of 

 a good fufpenfion. 



When the needle of the compafs difpofcs itfelf in the magnetical meridian, there is a cer- 

 tain line within the piece of fteel, which joins its two poles, and may be confidcred effec- 

 tively as the needle itfelf. But as this line is not vifible, the admeafurements of pofition 

 muft be made with regard to fome marks on the extremity of the needle ; which marks will 

 be truly placed when the needle is found to occupy the fame pofition with refpecSl to a fixed 

 point, upon being rcverfed, fo that the lower fide fliall become the upper. If the magneti- 

 cal power had been found on ex;iericnce to occupy the fame parts of the needle with pro- 

 portional intenfity, during its decay, to that it pofleffed immediately after the touch, thefe 

 marks, once made, would continue to fhcw the true magnetical points as long as the needle 

 poffefled any dii:e£tive force. Hut it is well known that foft fteel lofes its magnetifm 

 fooner than hard ; and confcquently it may be inferred that, unlefs both fides of a needle 

 were equally hard, the, magnetic power would deviate in procefs of time towards the harder 

 Cde. Thefe confiderimons lead to an obvious method of diminiftiing fuch growing error. 

 It confiits in making the needle flat and thin, and fufpenning it with its edge, and not its 

 flat fide, uppcrmoft, as is more commonly done. 



The needle is ufually fupported on a fteel point, which occupies the axis of a cylindrical 

 box called the compafs box. For this purpofe there is formed in the needle itfelf a cap or 

 hollow conical centre of brafs, fteel, or hard ftone, which is applied over the point. The 

 tendency of the needle to be difturbed by agitation will greatly depend upon the pofition of 



the 



