132 Messrs. Morrow and Watkin on an Interference 



one at right angles to the other two, both extremities of the 

 diameter of a ring being thus under observation. When in 

 perfect adjustment, the centre of the rings remains midway 

 between the two cross-wires. 



The displacement is produced by turning the lever V which 

 is pivoted independently of the remainder of the apparatus, so 

 that vertical pressure on it is not transmitted to I and b. 

 The sleeve c is kept in contact with two worked surfaces on 

 g by the piece h pressed against it by the springs e e, and 

 at the same time its contact with the point of s is ensured by 

 the springs d d. 



To attach an extensometer the piece 6 can be lowered in 

 its sleeve by releasing the set-screw ; the ends of the instru- 

 ment can then be passed between p and q. Where this is 

 not possible, the springs d are turned outwards so that the 

 tube b can be removed altogether. 



When the extensometer is one to be used in a horizontal 

 position, the calibrator can be laid on its side resting on the 

 legs n n. 



The sodium light is brought to a focus at p by a condensing 

 lens. It is obtained by placing a piece of platinum foil in the 

 flame of a Bunsen burner, and allowing a solution of common 

 salt to trickle very slowly down the foil. 



Other methods which have been tried do not give a 

 sufficiently intense light for a long period. This one, when 

 properly arranged, burns continuously with little or no 

 attention. 



Each new ring corresponds to a distance of 29'46 x 10~ 6 

 centimetres. / 



Test of an Extensometer. 



The points to be examined in testing an extensometer are — 

 the value of the magnification constant ; the action of the 

 instrument at starting and reversing (in order to discover 

 whether there is any ' lost motion ') ; and lastly — within what 

 limits the readings obtained are proportional to the displace- 

 ment and what correction is to be applied when this is not 

 the case. 



The following tests will serve as examples : — 



The extensometer was of the differential mirror type 

 arranged so that a displacement of 3 x 10 — 6 centimetres could 

 easily be detected. 



In the first place simultaneous readings of the extensometer 

 and calibrator were taken for successive displacements in the 

 same direction. 



The test commenced with the extremities of an outside 



