VOL. XLVI.] PHILOSOPHICAL TRANSACTIONS. 65 



a whole point on either side. One of them, which generally varied 6 or 7 de- 

 grees to the east, being drawn the same way, would stand at l6°. 



All these irregularities are owing to the structure of the needle : for the wires, 

 of which it is composed, are only hardened at the ends ; and that is done by 

 making the ends red-hot, and quenching them in water : if all these ends are not 

 equally hard, or if one end be hardened higher up than the other, when they 

 come to be put together, in fixing them to the card, that end which is hardest 

 will destroy much of the virtue of the other ; by which means the hardest end 

 will have most power in directing the card, and must consequently make it vary 

 towards its own direction. If you retouch these wires when fixed to the card, 

 the error will still remain ; for that wire which is best hardened will always be- 

 come the strongest. ; 



The wires being disposed in the form of a lozenge is the reason why these' 

 cards had so little force, that they might be made to stand at the distance of se- 

 veral degrees, on either side the point from which they were drawn. For all 

 magnetical bodies receive an additional strength by being placed in the direction 

 of the earth's magnetism, and act proportionably less vigorously when turned 

 out of it. So that, when such needles are drawn aside from their true point, 2 

 of the parallel sides of the lozenge will conspire more directly than before with 

 the earth's magnetism ; and the other 2 will be less in that direction : by which 

 means the first 2 sides will very much impede its return ; and the latter 2 will 

 have that impediment to overcome, as well as the friction, by their own force 

 alone. 



The needles that are used on board the men of war, and some of the larger 

 trading ships, are made of one piece of steel, of a spring temper, and are broad 

 towards the ends, but tapering towards the middle, where a hole is made to re- 

 ceive the cap. At the ends they terminate in an angle greater or less, according 

 to the skill or fancy of the workman. Now, though the worst of these are infi- 

 nitely preferable to those of wire, yet the best of them are far fi-om being perfect. 

 Every needle of this form has 6 poles instead of 2. There h one at each end, 2 

 vv'here it becomes tapering, and 2 at the hole in the middle. This is owing to 

 their shape ; for the middle part being very slender, it has not substance enough 

 to conduct the magnetic stream quite through from one end to the other. All 

 these poles appear very distinctly, when examined with a glass that is sprinkled 

 over with magnetic sand. Yet this circumstance does not hinder the needle 

 from pointing true; but as it has less force to move the card, than when the 

 magnetic stream moves in large curves from one end to the other, it is certainly 

 an imperfection. 



Two needles, that were quite straight, and square at the ends, were found to 



VOL. X. K (U ..jr.v i|£i) oji4^„ oni 



