Hackett — The Twist and Magnetisation of a Steel Tube. 417 



overcome hysteresis with respect to the intensity of the applied field, but also 

 with respect to its direction. It is cpiite obvious that the spiral field S will 

 not produce similar intensities of magnetization if produced by circular and 

 magnetic fields applied successively. Theoretically the spiral field should be 

 allowed to grow to its full value, keeping the pitch- angle constant, that is, the 

 ratio Ji/F should be constant. This is feasible by combining the circuits 

 producing the circular and magnetic fields so that they are both made 

 simultaneously. This method has been used in the later stages of the 

 investigation, but in practice it is hot so simple as another process which 

 was first adopted, and which is described in connexion with the actual 

 observations. 



Magnetizing Solenoid. — These experiments were rendered possible by the 

 use of a specially large magnetizing solenoid designed by Professor Brown, 

 for the use of which the author of this paper is greatly indebted to him. Its 

 length over all is 121 cm., and the magnetizing coil is divided into six sections, 

 each of which is wound so as to give a magnetizing field of 12-21 c.g.s. units 

 per ampere. The field inside the coil is uniform within 10 cm. of the ends, 

 giving a uniform field over 100 cm. The external diameter of the windings 

 is 7"62 cm. and the internal diameter is 2-54 cm. The internal space is pro- 

 vided with a jacket for heating or cooling as may be required. The whole is 

 mounted vertically in a strong wooden frame, 170 cm. high and 58 cm. wide. 

 Each end of the steel tube was brazed to a short length of brass tubing which 

 reduced its free length to 91 em. The upper length was clamped in a nut 

 rigidly attached to the wooden frame in which the solenoid was mounted, 

 so that the steel tube was entirely within the uniform field of the solenoid. 

 The lower length projecting below the solenoid carried the mirror reflecting 

 the beam of light whose deflection measured the twist. An insulated copper 

 wire passed down the steel tube, and was fixed centrally by ebonite plugs, 

 to carry the current which produced the circular field. 



Measurement of twist. — The difficulty of getting a beam of light to give a 

 clear, definite image on a scale at a sufficient distance away, to measure the 

 small deflections to be expected, was surmounted by utilizing the collimator 

 of a spectroscope. The image of a Nernst filament was focussed on the slit, 

 and the collimator adjusted to give an image on a glass scale at a distance of 

 1325 cm. "When the slit was narrowed sufficiently, the image was traversed 

 by diffraction bands approximately 04 mm. in width. The position of one of 

 these bands was read by a low-power ocular, and readings could be made with 

 ease to 04 mm., which corresponds to a twist of 3'77 micro-radians. 



3x2 



