by Magnetizing Iron by Frictional Electricity. 227 



tery, and if there is in a 6 a bundle of iron wires or a massive 

 bar of iron or nickel, when the current ceases all the actions 

 of the current in a /3 are increased ; the current which is ob- 

 served after the equilibrium has been destroyed must therefore 

 pass from « to /3. [The thermic, physiological, and galvano- 

 metrical effects were examined, also the power of magnetizing 

 steel and soft iron.] The separation of the iron bar into 

 wires, or, in other words, the destruction of the current efg h, 

 has no visible influence on the strength of the current as mea- 

 sured by the galvanometer and by its power of magnetizino- 

 soft iron, but it increases the physiological effect and the 

 power of magnetizing steel to an extraordinary degree, as I 

 have shown at length in a former paper. (Reports of the 

 Berlin Academy, 1839, p. 163, and Poggendorff's Annals, 

 vol. xlix. p. 72.) If a bar of unmagnetic metal, or a simple 

 tube (that is, e/g h without s n), be inserted into a b, then the 

 galvanometric effect of the current « /3 is not altered, but phy- 

 siological effect and the power of magnetizing steel are dimi- 

 nished. 



2. If the primary current is that of a self-discharging Ley- 

 den jar, then 



a. The thermic effect of the current « /3 is weakened both 

 by the electro-magnet 5 7i alone, and by the wire efgJi, and 

 also by both together; for the heating effects are diminished 

 by the insertion of bundles of wires or of bars of either mag- 

 netic or unmagnetic metals into a b. 



b. The physiological effect of the current a /3 is weakened 

 by the wire efgh either with or without s n, but strengthened 

 by this latter alone ; for it is diminished by a massive bar of 

 iron, and by a bar or tube of an unmagnetic metal, but con- 

 siderably increased by a bundle of iron wires. 



c. The electroscopic effect is just the same, for the current, 

 which may be tested by means of the condenser and resinous 

 figures, is found to proceed from (3 to « if there is in a i a bar 

 of iron or of an unmagnetic metal, but from « to /3 when it is 

 a bundle of iron wires or a bar of nickel. In this latter badly- 

 conducting metal the excited curveni efgh is too weak to 

 overcome the action of 5 n. 



d. Tiie power of magnetizing steel is, on the contrary, in- 

 creased by sn alone, and by sn surrounded with efgh, but 

 diminished by efgh alone; for the polarity of a steel needle 

 magnetized by the stream induced in « /3 shows that its direc- 

 tion is from a to /3 if a bar or bundle of iron or nickel wires 

 has been employed, but from /3 to a if a bar of an unmagnetic 

 metal has been used. 



Q2 



