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



The iron bar, with its stirrup, was pendent from the end of a 

 balanced lever placed over the pole of the electro-magnet, while the 

 arm of the lever, on the other side of the fulcrum, was weighted with 

 a sliding weight, or with variable weights, to balance the attractive 

 force of the iron when in contact with the electro-magnet. 



For the measurement of smaller magnetic forces, a special balance 

 was constructed which, besides balancing forces up to 15 Ibs., would 

 turn with a weight of less than half a grain. 



Preliminary experiments were made upon the bar by placing it, 

 when cold, in the direction of the dip, with one of its ends at a 

 definite distance from the magnetometer, and in the same horizontal 

 plane. When in this position, the magnetic force of the iron bar, aug- 

 mented by the earth's magnetism, produced a deflection of the needle 

 from the magnetic meridian of 20. 



The bar was then heated to bright redness, and replaced in the 

 same position as before, when all the phenomena described by 

 Barlow were reproduced. The heated bar had no perceptible action 

 on the magnetometer, bat on cooling down to a less red heat the 

 magnetic action of the iron began to manifest itself, gradually at- 

 first, and then very rapidly, till the deflection of the needle, which 

 was 20 with the cold iron, now advanced to 43, thereby showing an 

 increase of the magnetic power of the iron at this temperature. On 

 the further cooling of the bar, the magnetic action of the iron 

 gradually diminished, till the same deflection of the needle was 

 obtained as at the commencement of the experiments. 



The increase in the magnetism of the bar, as shown by this experi- 

 ment, although greatly augmented by the earth's magnetism, was, 

 however, very feeble ; for when the bar was placed horizontally at the 

 same distance from the needle the deflection was only increased 5 by 

 heating it to the temperature most favourable to its magnetisation ; 

 and no increase of magnetism could be perceived when a small 

 electro-magnet, however feebly excited, was brought into direct 

 contact with the heated iron. 



The bar was again heated to incipient whiteness, and placed over the 

 pole of the electro-magnet. As the cooling proceeded, observations 

 were taken of the intervals of time required before the magnetic 

 force was sufficient to cause the bar to adhere to the pole of the 

 electro- magnet for each definite increase of weight on the lever. 



The colour, and several shades of colour, of the heated iron above 

 visible redness, progressing towards the orange and yellow, are 

 expressed in wave-lengths of well-known spectral lines of the 

 alkaline and alkaline earth metals in the arc spectrum, as observed 

 through a direct-vision spectroscope of five prisms. Below these 

 temperatures, I have selected the melting points of zinc (442 C.) 

 and tin (230 C.), small fragments of which metals could be dropped 



