in different Latitudes. 103 



the doulle touch*. Indeed, in the application of tlii« process, 

 the needles, by the influence of the extreme bars, are brought 

 instantaneously to a much stronger degree of magnetism than 

 that which they could preserve when they are left to themselves; . 

 so that, if their internal constitution remain the same, the de- 

 gree of magnetism in which they are fixed must also remain 

 the same, or at least can only experience that degree of altera- 

 tion which arises from a change of intensity in the magnetic 

 force of the globe. This method might be rendered much more 

 certain bv keeping thus a stated number of well tried needles, 

 the separate effects of which might be ascertained at any pre- 

 vious stated epoch. For if, in trying them again at any other 

 time, it should be found that they have preserved among them 

 their original properties, we might conclude with certainty that 

 they have not teen altered in their constitution, and consequently 

 the observation of their absolute energies would determine the 

 real state of the magnetic force. 



But the adoption of this method requires the preservation of 

 the needles, aud the assurance of their identity. This care 

 might be dispensed with, if we could find a metho<l of fabricating 

 two needles capable of being accurately compared together at 

 anv time. For this purpose we must not think of employing 

 steel, which, being a mixture of carljon and iron, is necessarily 

 variable in its proportions. But this defect might be supplied, 

 if we could procure some wire of perfectly jiure soft iron by che- 

 snical means. For, according to the experiments of Coulomb, 

 which we have already mentioned, twisting gives such a degree 

 of density and elasticity to iron, that it takes magnetism nearly 

 as well as steel, and retains it equally as long : it is, therefore, 

 only necessary to regulate the quantity of twisting. Now, this' 

 is easily done by taking wires of an assigned length and size, 

 and measuring tlve number of turns which they are twisted by 

 means of a micrometer. Each of these wires should then be mag- 

 wetized to saturation, and a certain numl)er of them should be col- 

 lected together, so as to form a bundle, the directing force of which 

 should be measured, either by a magnetic Ijalauce or by the me- 

 thod of oscillations. All the'difficulty is thus reduced to the pro- 

 curing of pure iron, and this difficulty belongs solely to chemistry. 



If in any given place a piece of soft iron, such as a key tor 

 example, be put in the direction of the magnetic dip; tlie key 

 presently acquires two poles, which act on the tnagnetic needle 

 in contrary directions. Turn the key about, and its poles will 

 change, and the same effect may be reproduced. 



• See Supfjlcjnent to the 3(1 ulition of the Fnn/.Jirit. nrt. Ma^uithm, 

 No. .'>9; and lor a list of :ill iwa lust writers oti tins Mihject, Me tlie end 

 of tliiit article, No. 89; or Dr. Thomas Young's verv v;iliial)!c Tresiti^o of 

 Natuial Fliiiosophy, vol. ii. j)agc 437. 



