Elongation produced in Soft Iron by Magnetism, 499 



otherwise result from so long a bar, and destroy the definition 

 of the fringes. 



A telescope T was used to observe the displacement of the 

 fringes due to the forward or the backward motion of the 

 mirror m' . A motion of one wave-length produces a dis- 

 placement of two fringes ; so by observing in the telescope the 

 number of fringes-displacement relative to the cross wires, it 

 is an easy matter, knowing the wave-length of sodium light, 

 to compute the actual expansion or contraction of the bar. 



I had also at my disposal a single-thread magnetometer, 

 the magnet being suspended by an exceedingly fine quartz 

 fibre. 



There were six different bars used in these experiments. 

 We will hereafter refer to them by number. 



No. I. w r as 70 cm. long, 0*952 cm. in diameter. 



No. II. was 70 cm. long, 1*032 cm. in external and 0*714 cm. 

 in internal diameter. [ 



No. III. was 70 cm. long, 0*635 cm. in diameter. 

 *No. IV. „ 43*5 cm. „ 0*635 cm. „ 



No. V. „ 19*5 cm. „ 0*635 cm. „ 



No. VI. „ 70 cm. „ 0*752 cm. „ 



With the exception of bar II., soft unannealed Norway iron 

 was used. No. II. was apiece of ordinary wrought-iron gas- 

 tubing. Nos. IV. and V. were made from No. III. by cutting 

 off one end. No. VI. was made from No. I. by turning it 

 down in diameter. 



In the case of the short bars, brass rods were soldered to 

 both ends. To make certain that brass w r as not at all affected 

 by magnetic influence, I tested a brass rod and found no 

 effect whatever. 



With this instrument I found it an easy matter to measure 

 to a tenth of a fringe, which means about 2 00 q 0oo of an inch. 



In making; these observations the only difficulty which I 

 found at all troublesome were certain effects due to hysteresis. 

 The expansion due to the heating of the coil being slow in 

 its action while the magnetic expansion was rapid, the two 

 could be easily distinguished from each other. The object of 

 using the glass tubing for my solenoid being to retard as 

 much as possible the heat action, the heaviest currents used 

 could remain on at least 5 seconds without showing any effect, 

 then the bar would slowly begin to expand. 



Mr. Shelford Bidwell, in mentioning the difficulties which 

 he encountered, states that he found it absolutely necessary to 

 centre the bar in the coil. He finally adopted the method of 

 suspending the coil on one end of his bar, so that it moved 

 with the bar. In this way the effect was entirely obviated. 



2 L2 



