MAGNETISM. 



in the effect of the force applied to 

 them. This circumstance has escaped 

 the attention of the most skilful ob- 

 servers, although it is of sufficient im- 

 portance to destroy all confidence in the 

 comparison of experiments made in 

 places where the dip is very different *. 

 (339.) We have seen the method by 

 which the amount of the dip may be 

 determined from the comparative inten- 

 sities of magnetic force in the plane of 

 the magnetic meridian, and in the hori- 

 zontal plane. It is easy to see how, by 

 reversing the process, we may arrive at 

 the knowledge of the force of terrestrial 

 magnetism in the magnetic meridian, 

 from observations of the intensity in the 

 horizontal plane, when the dip is pre- 

 viously known ; for we have only to 

 augment the former in the proportion 

 of the cosine of the dip to the radius : 

 or, what comes to the same thing, to 

 multiply it by the secant of the dip. 

 For since 



it Mows that 



cos. 



h. sec. 



(340.) The method of oscillations is 

 ttot the only mode of determining the 

 intensity of magnetic forces ; for the 

 same object may be attained by employ- 

 ing the balance of torsion : but in gene- 

 ral the former method is, on many ac- 

 counts, preferable. 



(341.) It would conduce much to the 

 future advancement of the science of 

 Magnetism, if the absolute intensity of 

 the terrestrial magnetic forces could be 

 ascertained by a standard equally deter- 

 minate as those by which we estimate 

 the atmospheric pressure, or the tem- 

 perature of different climates; for by 

 repeating the same process of observa- 

 tion, in the course of successive centuries, 

 we might learn whether the intensity 

 of these forces experienced any variation 

 similar to what we know takes place 

 in their direction. 



(342.) The method which first sug- 

 gests itself, would be to observe the 

 variation, the dip, and the intensity, by 

 means of three needles, respectively ap- 

 propriated to these objects, and carefully 

 preserved, with a view to a repetition of 

 the same experiments at distant periods 

 of time. As it is possible that their 

 magnetic power may become impaired 



* Poniiiet, Elgmens de Physique Experimental 

 torn. i. \\ 4S2, 



by these long intervals of time, it would 

 probably be necessary to magnetize 

 them afresh, employing for that purpose 

 the most efficacious methods, so as to 

 induce a degree of magnetism which 

 shall at first be considerably greater 

 than what they are capable of retaining 

 permanently. When left for a certain 

 time to themselves, they will then return 

 to their natural state of saturation, and 

 in this state they may be subjected to 

 the experiments in question. Our as- 

 surance of their attaining this determi- 

 nate degree of magnetism may be in- 

 creased, by preserving a great number 

 of needles; brought to this state, and 

 noting their respective powers ; which 

 records may afterwards be compared 

 with similar observations made at a 

 future period; for if they should be 

 found to have preserved among each 

 other the same relative degree of strength, 

 we may confidently conclude that no 

 alteration has taken place in their mag- 

 netic constitution, and that they are well 

 adapted to the determination of the ab- 

 solute force of terrestrial magnetism. 



(343.) If a method could be dis- 

 covered of obtaining a material for the 

 making of magnets of perfectly uniform 

 composition and qualities, there would 

 be no necessity to employ in this com- 

 parison the same identical magnets at 

 the several times of observation. Steel 

 would evidently be unfit for the purpose, 

 its composition being necessarily varia- 

 ble. Nothing would answer so well as 

 iron alone, wliich might be obtained in 

 a state of perfect purity by proper che- 

 mical processes ; for Coulomb found 

 that pure iron, however soft, may be 

 rendered nearly as retentive of magnet- 

 ism as steel, by merely twisting it. All 

 that would then be necessary, would be 

 to regulate the degree of twist, so that 

 it might be constantly the same for dif- 

 ferent pieces of iron : this might easily 

 be accomplished, by taking them exactly 

 of the same dimensions, and measuring 

 the number of turns in the twist given 

 to them. Each of the bars might then 

 be magnetized to saturation, and em- 

 ployed either separately or in defined 

 combinations, and their magnetic forces 

 examined, both by the torsion balance, 

 and by the method of oscillations*. 



(344.) Poisson has also attempted to 

 resolve this problem by considerations 

 of a mathematical nature, for which we 

 must refer to his memoir, read to the 



Biot, Traite" de Physique, torn, 



