1891.] upon the Magnetisation of Iron and other Substances. 113 



The results also show a rapid increase of the magnetic power of the 

 bar from X 6496 to A, 6705. This increase was attributed, in the first 

 instance, to an error of observation, but on repeating the experiment 

 similar results were obtained. The rapid increment of magnetic 

 force in the interval of cooling between these spectral lines may, there- 

 fore, be regarded as a real phenomenon. 



In all the experiments which have hitherto been made, where an 

 increase of the magnetic power of iron with increase of temperature 

 has been observed, it does not appear to have been suspected that the 

 mass of the iron, in relation to the magnetising forces employed, 

 might be an important factor in the results obtained, and that 

 small magnetising forces might only penetrate to a small depth 

 below the surface of the iron, when cold, till the more central por- 

 tions of the mass were brought into action by increase of tempera- 

 ture. Several magnetic properties of iron and steel, however, point 

 to the probability of this action of weak magnetising forces. Coulomb 

 found that the magnetism of similar steel bars did not increase in the 

 ratio of their number when laid together, from which it was inferred 

 that the magnetism diminished .from the surface to the centre of the 

 bars. Joule has shown that a hollow electro-magnet has greater 

 attractive force than a solid one of the same sectional area, with a 

 small magnetising current.* It -is also well known that the dis- 

 tribution of magnetism on the polar surfaces of electro-magnets is 

 much greater at the circumference than it is at the centre. 



That small magnetising forces penetrate but a small depth into a 

 mass of iron was shown by heating one of the bars to redness, and 

 sprinkling over its surface ferrocyanide of potassium, in powder, 

 before plunging the heated bar into cold water. The conversion of 

 the surface of the iron into steel by this well known process was 

 sufficient to reduce the deflection of the needle from 20 to 15, when 

 the bar was placed in the direction of the dip. 



In addition to the evidence adduced of the surface action of small 

 magnetising forces on a mass of iron, it further appeared to me that 

 as the time and limit of magnetisation of iron vary with the mass, for 

 constant magnetising currents, or that the time and limit of magneti- 

 sation are constant when the magnetising current and mass vary in 

 proportion, so it also appeared to me that, when the mass of iron and 

 magnetising force were proportional, the diminution of the magnetic 

 power of iron with increase of temperature would be constant for 

 small, as well as for large, magnetising forces. 



That ihe increase of the magnetic power of the heated bar, as 

 shown by the magnetometer, was caused by the large mass of the 

 iron in relation to the magnetising force of the needle was shown 

 by the following experiments : 



* ' Annals of Electricity,' 1840, vol. 4, p. 60. 

 VOL. L. 1 



