MAGNETISM. Q*^ 



^vlll not break the conne6tion formed by Its prefence between 

 the bar of foft iron and the wire; for the latter will remain 

 attached to the end of the former, by the extremity which 

 was firft brought into conta(51 with the iron ; if the piece of 

 wire be removed from the end of the bar, the magnetic 

 connexion may be revived by replacing it immediately. The 

 fame thing will happen if the wire be expeditioufly transferred 

 from the firft bar to another rod of foft iron ; but it lofes its 

 magnetifm in the fpace of two or three feconds when kept 

 at a diftance from all ferruginous bodies which are capable of 

 attrading it, and of being attraQed by it. Thefe fafts prove 

 wire of number 32 to be a magnet, the virtue of which is 

 conditional, becaufe its permanency depends on the prefence 

 of foft iron, and perhaps on no other circumftance ; for the 

 experiment may be repeated with fuccefs upon rufty wire of 

 the fame fize, or on pieces which have been made red hot in 

 the flame of a candle, or furrounded by fand in a crucible, 

 in which lituation they will cool much more gradually tfcan 

 when drawn fingly through a flame. 



Exp, 3. This capacity of iron wire to preferve the magr ^'"^ot cq^. 

 nctifm imparted to it, as long as it remains in contaft with a ditional magnet. 

 bar of the fame metal, is a property confined to certain fizes ; 

 for let the firft experiment be repeated with a fmall piece of 

 numbers 18 or 17, not equal to half a grain in weight, and 

 juft as it comes from the hand of the workman, this piece will 

 perform the part of a particle of foft hammered iron, that is 

 it will drop from the end of the bar, to which it has been 

 attached by the application of a magnet, to the oppofite 

 extremity, as foon as the magnetic influence ceafes to aft 

 upon it : confequently the mere operation of drawing foft 

 iron into wire, by forcing it through a conical holje too narrow 

 for its prefent diameter, will not convert it into a conditional 

 magnet. 



Amongft other experiments relating to the fubjed, I took Theloweft fize 

 the trouble to examine the quality of every fize from 32 to 2 1 , ^i^^ong, ^ag^ne". 

 both inclufive; the 11 fmalleft wires, the extrearas of which tifm afcertainefl. 

 were 32 and 22, were all conditional magnets; that is, they 

 all adhered to the bar of foft iron, to which tJiey had been 

 previoufly attached, after the removal of the magnet. 

 Number 23 fupported feven grains of lead including its own 

 weight, without .the affiftance of the magnet; No. 2>i, 6f 

 aearJy j No. 32, 4.| j No. 22, no moye tha» two grains. 



As' 



