82 Method of exciting Mngiietlfm. 



augmented by the aclion of the two.fupenor magtieis, which' concur In magi.^tizing all the 

 particles of the neeJle in the fame dh-eftioh. ' - 



As the needle n s, placed between the two large magnets ih the preceding operation, 

 acquires bythe joint adtion of the four bars a degree of polarity which is more than it can 

 preferve when feparated from them, it follows that at the mofheiit of this feparation the 

 needle lofes part of the magnetifm it derived from thofe forces, and that its magnctifm di- 

 miniliies Until the' inagnetic a£lion of the whole needle on each of its parts is in equilibrio 

 ■with'the coercive force. Hence, upon feparating the needle from- the magnets, it is found 

 to be faturated with magnetifirt. 



I have found likewife, that in this method of magnetizing, there is a greater certainty of ; 

 giving to both furfaces of needles, intended to determine the magnetic meridian, an equal 

 degree of magnetifm J a circumftance deferving of the greatefl attention in the conftruc- 

 tion of compafles, if the needle be fufpended with its broadefl furface parallel to the ho- 

 rizon. 



The ConfruBion of Artificial Alagrit'fs. 



I TAKE fig. 4. thirty bars of (leel hardened and tempered to the temper of a fpring * 5 or 

 6 Uneshroadi.a or 3 lines thick, and 36 inches long. The blades of fencing foils, fuch as 

 are found in the fliops, make pretty good magnets. Englifli .fheet fteel f (la tole d'acicr 

 d'Angleterre) cut uito pieces one inch wide, hardened and lowered to fpring temper, is pre- 

 ferable. When each compound magnet is to contain no more than 15 or 20 pounds of 

 fteel, it isfufficient to make the bars 30 or 36 inches long. 



I magnetize each "bar fingly, according to the method already defcribed. I then take two 

 reflangular parallelopipedons of very foft iron, v.'dl polilhed, 6 inches in length, between 20 

 and 24 lines broad, and 10 or 12 lines thickj : with thefe two parallelopipedons reprefented 

 fig. 4. at N andS, I form the armor of my inagnetby enveloping one extremity of each pa- 

 rallelopipedon with a ftratum of my magnetic bars, fo that the extremities of the parallflo- 

 pipedons may projeil beyond the extremities of the bars 20 or 24 lines, and the other part 

 may be enveloped by the ends of the fet of bars. On this firft layer of fteel bars of 3 or 4 lines 

 thick, I place a fecond which Is tliree inches (horter than the firft, fo that the firft projeds 

 beyond the fecond 18 lines on each fide. The whole is fecured at the ends by two binding 

 pieces of copper, which prefs the bars clofe tpgefher, and prevent the armor from efcaping. 



Ftg. 4. reprefents two artificial inagnets compofed according 'to the method juft de- 

 fcribed. N and S are the extremities of the two Iron parallelopipedons. The two other 

 extremities are inclofed by the bars, liach magnet thus compounded is folldly conne£ted 



• By other experiments, the author has afcertained that the mag.ietifin acquired by fteel is leaft when it ij 

 either ablbhalely hard or annealed by a white heat, and grcatcft whcw'it has licen hardened and annealed by a 

 very obfcure red heat ; from which middle term it dimipiflies eithcs Wayv,Avlmtlicii the bar be harder or 

 Ibfter. N. , : ■ , ' . v.;,r i::: '' ;. : 



f We have two fons at leaft in thel-ondon market. The commoa fort is fol^ reoil at.fid. or 7d. per Ih. 



the finer, under the name of caft-fteet (vvhiLh I fuppofe it to be) is ("old at one fiiilliiig per lb. and deferves its 



price. I have noc been able to procure fiicet fteel thicker than i-f:th of an inch. N. ' 



I All the weights and meafurcs in this paper are French ; but as extreme prcciflon is no where implied, it 



was unnecefl'ary to reduce them. N. . 



3 together 



