510 Mr. H. F. Newall on the 



with sufficient accuracy for such purposes as the determina- 

 tion of coils in compound winding. But the probable charac- 

 teristic may be determined in the way here described, with 

 sufficient accuracy for a number of useful purposes. It is 

 based on the facts: — (1) When a machine is working at its 

 best permanent output, its iron magnetic resistance plus the 

 air magnetic resistance of the clearance is equal to the air 

 magnetic resistance of the space on the outside of the arma- 

 ture occupied by winding. (2) At the beginning the air- 

 resistance is alone of any importance. 



LXII. On the Recalescence of Steel. By H. F. Newall, 

 M.A., Demonstrator in Experimental Physics, Cambridge 

 University* . 



IN a Note published in the Philosophical Magazine for 

 November 1887, I stated the results of a number of 

 experiments made with a view to investigate the recalescence 

 of steel. I propose in the present note to give an account of 

 the experiments on which I base those results. 



I. Reglow, or recalescence, is not due to chemical action at 

 the surface of the steel, nor is it due to occlusion of gases. 



A steel wire with copper terminals was put in a glass tube, 

 the terminals passing through indiarubber stoppers, which 

 closed the tube tightly at its two ends. The tube was ex- 

 hausted by means of a mercury-pump till the air-pressure was 

 a very small fraction of a millimetre. The steel wire was 

 heated again and again by the passage of an electric current 

 from storage-cells. Pressure rose slightly after the first three 

 or four heatings (probably because of the setting free at high 

 temperatures of the air-layers condensed on Avire and tube 

 surfaces), and was reduced by further pumping. Both dark- 

 ening and reglow were observed with undiminished distinct- 

 ness after fourteen heatings, and the process was not repeated. 

 Nitrogen was admitted to the tube and produced no effect 

 that could be detected on the brilliancy of reglow. 



II. Reglow is not due to difference in conductivitv at 

 different temperatures, as suggested by Forbes. 



A wire ('5 millim. diameter) was hammered flat and still 

 showed reglow. A thin steel plate, hammered till its thick- 

 ness was less than -1 millim., as measured by a screw-gauge, 

 still showed reglow. Either we must believe that there can 



* Abstracted from a paper read before the Cambridge Philosophical 

 Society, January 29, 1886. Communicated by the Author. 



