On the Magnetisation Limit of Iron. 



291 



the experiments show that the viscosity of the film of water at the 

 •surface has a much greater effect, the heat transmitted in this case 

 being sensibly inversely proportional to the mean viscosity of the 

 iilm. Experiments on the jacket water confirm this result, so that 

 equation (2) becomes : — 



*L= ^ *- T - P. ^ 



where P iM is the value of P, calculated from the mean value of 



-i(T + t). 



According to the experimental results, the value of the coefficient 

 is in all cases about 5 per cent, less than the value of V m thus 

 calculated. 



64 On the Magnetisation Limit of Iron." By Henry Wilde, 

 F.R.S. Received April 3,— Read May 13, 1897. 



In my paper " On the Influence of Temperature upon the Magneti- 

 sation of Iron," I described a new method of determining the mag- 

 netisation limit of magnetic substances, by which, with a single pole 

 of an electro-magnet, a more exalted degree of magnetisation was 

 indicated, as measured by the force of traction, than had previously 

 been attained.* 



The magnetisation limit of iron, as deduced from my experiments, 

 was 381 lbs. per sq. inch of section, and it appeared to me at the 

 time that the extreme limit was well within 400 lbs. per square inch. 



I have recently had occasion to repeat these experiments with 

 other specimers of iron of different lengths, and have increased the 

 magnetisation limit sufficiently high to place the result on record. 



Definite lengths of annealed charcoal-iron wire of circular section, 

 0*57 inch diameter, as measured by a micrometer screw-gauge, were 

 cut from the same piece of wire, and mounted for suspension over 

 the electro-magnet in the manner described in my former paper. 



The specimens of iron were submitted in succession to the action 

 of the electro-magnet excited by a current of 40 amperes, and the 

 force of traction was measured by the weight required to detach the 

 iron wires from the magnet, with the following results : — 



* < Roy. Soc. Proc.,' 1891, toI. 50. 



